Rack Mounting FAQ

19 inch racks of a storage array
This text, mainly written in the form of a technical terms glossary, is meant as quick-start guide on rack mounted (computer) equipment as commonly found in server-rooms and data-centers. Short paragraphs per term summarize often complex matter to provide you with an overview of the field, a first primer on this fascinating topic. Get a grip on standards, industry terms, common abbreviations, rack dimensions, rack materials and more. It is also an introductory text on a number of metal fabrication techniques, on sheet metal forming, cutting, welding and coating. While racks are used in scientific, audio/video and very much in IT and ICT (Information and Communications Technology) applications - all of which commonly perceived as "white glove" fields, this basic text also shares how they are designed and made in an industrial hands-on process. Micropolis believes in the advancement of scientific and engineering excellence through support of education, of research and by sharing knowledge. So, let's try that and dive right in.

10-inch rack (10" rack)

While it is common to use one 19" rack unit for two devices side-by-side, dedicated less-wide racks are finding their way into IT deployments in recent years. 10" racks usually are less deep, in comparison to common 19" server cabinets, following the overall smaller footprint. 10-inch rack dimensions are: exactly 10" (254.00mm) from side to side horizontally (front plate), including screwing area. Actual usable width is 8.75" (222.25mm) clear area in-between vertical screwing rails ("rack opening" or "aperture"). Both standards, 10" and 19" use rails of same width, 0.625" (15.875mm). Apart from saving space in commercial installations, 10-inch rack frames are deployed in military, in airborne or otherwise weight or space constrained scenarios. While dedicated 10-inch switches, 10-inch hubs and 10" computer enclosures are available, it is common in SoHo or business environments to install, not specifically 10", consumer routers or similarly small devices in 10-inch racks in a DIY fashion on 10-inch rack shelfs. Audio devices like microphone receivers, tone generators and effects processors use a 10"-similar half-width rack format, the 9.5" footprint, for decades. And although 9.5" half-rack devices fit a 10" rack perfectly, some vendors offer 10.5" racks, with a slightly wider 264mm/266mm to 271mm/275mm front panel width and 249mm or a little more hole to hole distance.

19-inch rack (19" rack)

is a standardized type of technical furniture, a frame or enclosure, of roughly 19" width for mounting multiple electronic equipment modules stacked over another on two parallel vertical rails. The "EIA 19-inch Standard Rack" (EIA-ECA 310E) is very common in scientific and computer / data-center environments. Actually, only the front plate of installed module measures 19 inches (482.60mm) wide, giving the name. Such a front plate is 19" (482.60mm) in total, including the screw terminal area, with 17.75" (17 3/4 inch, or 450.85mm) of clear area in-between vertical screwing rails (called "rack opening" or "aperture"), the actually usable space for mounted equipment between the mounting posts. 19" racks may be embodied as open frames, offering essentially vertical screwing rails only ("two-post racks", popular in telco applications) or as open or closed cabinets (usually 24" wide, 609.6mm), with optional additional internal width outside the rail area for more elaborate cabling.

Rack Flanges

Each EIA 310 compliant rack features at least two vertical beams (vertical "mounting flanges", "mounting posts", "uprights" or "rack rails"). These two posts are spaced 17.75" (17 3/4 inch, or 450.85mm) apart, building a "rack opening" and this is the space where installed equipment is residing in or in-between, the "usable aperture opening" or "rack aperture". The front of rack flanges is expected to be 5/8 inch wide (0.625" or 15.875mm), but the important dimensions are the clear area in-between flanges and the distance of the rack holes on the flanges center to center. On racks where the mounting flanges are a recessed element, with double-L shaped mounting rails or when the rack flanges are part of or installed in a closed rack cabinet, the EIA specs expect the available width for front panels (the "panel opening") to be at least 19" plus 1/16 inch (or 19.0625" or 484.1875mm) toleranced at 1/32". Suggestions for the maximum width of installed equipment in-between the posts (including any slides, rails or similar mounting helpers) varies between a suggested 17.625" (447.675mm) in imperial unit specs and of course the absolute ceiling value imposed by the defined 17.75" (17 3/4 inch, or 450.85mm) minimum rack aperture. It may be noted here that soft converted metric specs gave 450mm at one point in time, a little less, as the minumum space between rack posts, but read a little below for a note on this soft conversion process.

Rack Holes

Rack mouting rails (vertical "mounting flanges") may have square unthreaded holes (9.5mm x 9.5mm), round threaded (tapped) holes or unthreaded round holes (clearance holes). Vertically, a standard rack flange has a repeating pattern of either two or three grouped holes. The pattern with a group of three repeating holes is called "universal spacing" and allows finer grained mounting positions, and the pattern with two repeating holes is called "alternative spacing" (or "wide spacing") as it simply omits the "center hole" of the group, resulting in holes being wider apart and enforcing a more fixed module layout within the rack. Some sources imply that this "alternative spacing" is the "older" hole pattern, but in fact these two hole spacing layouts are optional variants already part of the 1960s EIA 310 specs. When vertical mounting rails offer unthreaded (untapped) "clearance holes", EIA specs expect these to be 9/32 inches in diameter (or 0.28125", rounded to 0.281" or 7.1374mm), plus/minus a tolerance of 0.003" (0.0762mm). Metric soft converted specs expected bore holes to be 7.1mm (with 0.1 tolerance) in diameter. Horizontally, hole spacing is 18.312" (or 465.124mm) center to center, also with square holes, toleranced at 0.062". In EIA specs, imperial dimensions are given in decimals with a precision of three digits after the decimal point. Whether or not these decimals result from an underlying dimension of 18 5/16 inch with a tolerance of one fraction (1/16) inch is left to the reader. Whether this tiny uncertainty matters, under given tolerance, as well. One thing to note though is, that holes are not centered horizontally on the rack flange, but are slightly offsetted to rack center. Subtracting horizontal hole-to-hole distance of 465.12mm from the total outer width of 482.6mm gives an outer width of the rack flange of 8.74mm and subtracting the rack opening of 450.85mm from hole-to-hole distance gives 7.135mm inner width of the flange (both to hole center, of course). Regarding the wider outer rim of the flange, note that this is part of the "wiggle room" a rack must offer for front panels (read above in "Rack Flanges"): the rack flange is expected to offer a minimum total outer width ("panel opening") of 19" plus 1/16 inch (or 19.0625" or 484.1875mm) and 19.0315" or 483.4mm suggested elsewhere in EIA specs. A common layout is to have rack flanges 19mm wide (0.75") - with a rack opening of 17.75" this totals to 19.25". Rack holes here are offsetted 12mm (0.47") inwards from each outer side and 7mm (0.276") from the inner flange edge. Between vendors and flange type or material, frontal width ranges from 19mm to 21mm. Some rack types (mostly in the audio field) feature vertical rails made from Extruded Aluminum, where equipment may be mounted at arbitrary positions with a combination of screw and square nuts. Such flanges are often wider, 25mm. Some vendors offer non-standard 4-hole (per height unit) rack rails, where the center hole of a vertical three-hole group is omitted and replaced with two offsetted holes. This allows equipment to be mounted 0.5U vertically apart for better airflow between modules. Most racks however conform to EIA 310 and feature holes at a specified vertical spacing.

How tall is 1U?

One height unit (a "rack unit" ("RU"), measured in "U", 1U, 2U, 3U, etc.) is 1.75" (44.45mm) high. This height is based on the mounting hole distance. The original EIA 310 sepcifications used imperial units (inches and punds) and defined the grouped three mounting flange holes (always measured from the hole's center) as 0.625" (5/8 inch, or 15.875mm) apart. The next group of three holes then is spaced a little closer at 0.5" (1/2 inch, or 12.7mm). These patterns start from the top with a first half-step-0.5" (plus some extra space) distance of 0.312" measured from the top of the vertical rail. Distances 0.625" plus 0.625" plus 0.5" give a total of 1.75" height, which is 1U, "one rack unit", sometimes abbreviated as "RU" or height units ("Höheneinheiten", "HE" in German).

Metric "soft conversion" discrepancies

These 19" rack dimensions given in inches remained unchanged for many years. Then, during the 1990s, the U.S. government started a number of initiatives to make the country more competitive. As part of the 1988 Omnibus Trade and Competitiveness Act (public law 100-418), the imperial specs were revised in 1992 and the updated document EIA 310-D stated most dimensions in metric values, to better align with international practices and help users in metric countries. But reading the document can be puzzling, as many values are now metric but seem less precise. The explanation is that the actual specs weren't changed, but these metric documents use a process known as "soft metric conversion". In short, this means that imperial values weren't changed or adjusted, for example to align with their closest even metric numbers, with less digits or so. It only means that imperial values were converted to metric values that reflect, more or less, with a certain degree of inaccuracy, the underlying unchanged imperial values, like a rough guideline. These metric values are not fit for engineering application. For example, many values given in these soft converted documents expose a rounding error where values are multiplied, an error only explainable when the reader is aware of the fact that calculations are still done in imperial values, and only the final result is then converted to metric values. So, to sum all this up, 1U is not 1.752" or 44.50mm tall but is still 1.750" or 44.45mm. And hole spacing is not 0.626" or 1.252" but still 0.625" and 1.25".

Rack Panels

As can be seen in the illustration below, the 19" x 1.75" of a 1U rack slot are the defined maximum outer dimensions for a single rack position, and the maximum size of a 1U module or blank filler panel. While the screw positions inside this 1U envelope are hard defined, the actual outer size of a 1U unit is usually a little smaller than the available maximum and varies slightly between vendors. That said, as per EIA standard, the width of (blank or front) panels is expected to be exactly 19 inches (482.6mm, with a tolerance of 0.4mm), but the height of one panel (or the accumulated height of multiple panels) has to be "one fraction of an inch" or 1/32" or 0.03125 rounded to 0.031" less in height than the maximum available. The actual EIA definition was the "height tolerance" is "plus zero, but minus 0.031". This equals to 0.79375mm, rounded to 0.79mm or 0.8mm, giving modules or blank panels a little "wiggle room". To calculate the height of panels, use this formula: h = (1.750 * n - 0.031) inches = (44.45 * n - 0.79) millimeters, where n was the number of rack units occupied, and h gave the resulting panel height. So a standard conforming 1U panel is 1.71875" tall (rounded to 1.719" in EIA specs and equalling 43.6626mm) - inside the 1U envelope that is 44.45mm (1.750") tall/high. Some sources misinterpret EIA D from 1992 and later revisions and re-define 1U as being 1.752" or 44.50mm tall, but this also stems from a wrong understanding of the "soft conversion" process used in the specifications, as mentioned above. Panel thickness is left to the user to decide upon as needed for the application. Also, with larger, more rack units tall, equipment, the number of slots may be just four or more slots, to increase the load bearing capabilities of a module.

Rack Slots

Rack panel holes may be open slots or notches, opening all the way to the side edge of a panel, or closed slots resembling a wide hole. Both of these slot variants are standard as per EIA 310. The inward length of slots, from the side edge of the panel, the horizontal width, open or closed, is 0.531" (1/2" and a fraction, 1/32") at minimum, with open slots allowed to be 0.625" (5/8 inch) wide. Closed slots are expected to have a closed border, to the side edge of the panel, of 0.125" (or 1/8 inch). Both slot types, open and closed, are expected to be a minimum of 0.25" (1/4 inch) wide (tall), meaning the vertical height of the opening, with a maximum opening of 0.281" (1/4" + 1/32"). In metric units this means slots need to be between 13.4874mm and 15.875mm long/wide and 6.35mm to 7.14375mm tall. Closed slots must have a border to the panel side of 3.175mm. The position of the slots is mirrored along the horizontal center axis of a panel and aligns with the rack holes on the mounting flanges. Later metric EIA specs suggested open slots to be (soft converted/rounded) 13.5mm to 15.9mm long/wide and gave a little more leeway for the vertical height, suggesting slots to be 7.1mm tall (equalling 9/32 inches instead of 1/4) and toleranced with +/- 0.3mm, for 6.8mm to 7.4mm.
19 inch rack panel hole positions and size
Close-up of a 1U 19" rack front panel with slot positions and dimensions

Rack Screws

While the original EIA specs outline that pretapped mounting holes should be made to accept either a "number 10" (#10) fine thread machine screw (10-32 Class UNC-2B, sometimes 10-32 UNF) or a 12-24 Class UNC-2B as an alternative, the latter number 12 screw is more prevalent in US/imperial datacenters. A 12-24 screw has a slightly larger head and the thread is a little more coarse and this seems to help when mounting larger equipment. Audio equipment on the other hand is often fastened with a 10-32 screw. And to make matters worse, in metric racking environments, it is common to use a M6 machine screw. More recent revisions of EIA 310 prefer metric values and add M5x.8-6H and M6x1-1H screws to the officially defined screw sizes for compliant racks with M5 being the preferred thread size. One thing that this multitide of thread pitches clearly shows is that threads in general may be problematic. And not to mention that threads might get damaged by a wrong screw or repeated change of equipment. Of course, unthreaded "clearance holes" are part of EIA 310, but they require a second tool to hold the counter nut. This is why square holes (square punched "full holes") in racks were introduced. With a square hole an equally square nut can be locked or "latched" in place. A "cage nut" is a square nut wrapped in a small metal sheet (a "cage"), having two or four small spring clips which grip to the square hole. Once installed, the cage nut works like an adapter and can accept either M5, M6, 10-32 or 12-24 screws - whatever fits the installed equipment.
  • 10-32 Class UNC-2BF
    A "number ten" Unified National Coarse, 32 threads per inch (tpi) screw, compliant with the Unified Thread Standard (UTS). The designator "10" is an arbitrary label for the size of the screw, similar to Gauge, with no numerical meaning. The threaded part has a diameter of a little more than 3/16" (0.1900" or 4.8260mm). The V-profile of the thread is the same as for ISO metric screw threads, only the pitch is based on inch values. "2BF" is a tolerance class.
  • 10-32 UNF
    The number "10" Unified National Fine is the same as the UNC number ten screw, except that the thread pitch is smaller and the V-profile is less deep, leading to less travel per turn. Fine pitch screw variants are less common and it should be checked if a rack's threaded holes are really fine pitch.
  • 12-24 Class UNC-2B
    A "number twelve" screw is a little bigger than a #10 and a little more coarse. This screw has 24 turns per inch (tpi). The threaded part has a diameter of a little less than 7/32" (0.2160" or 5.4864mm).
  • M5x.8-6H
    Screws prefixed with "M" are from the ISO metric screw thread series, where the number indicates the screw's nominal outer diameter of its thread in millimeters. An M5 screw has a (major) diameter of 5.00mm. Note that through boreholes for this screw are expected to be a little wider according to EN 20273, 5.5mm. An M5x0.8 screw is the common type with a standard pitch of 0.8mm, or 0.8 threads per millimeter. "6H" is a (common) tolerance class, for V-profile and minor diameter. M5x.8 is the metric equivalent of an UNC screw.
  • M6x1-1H
    A metric screw with a nominal diameter of 6mm and the most common pitch of 1mm (one thread per millimeter). Compared to an UNC 12-24, this M6 screw's threaded part has a diameter of a little more than 7/32" and even 15/64" (0.2362205"). The EN 20273 standard through borehole for a M6 is a little wider, 6.6mm. With a pitch of 1mm, this screw is the "normal", "regular" or "ISO" pitched type, and the metric equivalent to an imperial UNC thread screw. Similar as with "6H", "1H" is a tolerance class.

Rack Positions

When installing equipment, it's necessary to find the right "starting position" so that 1U, 2U etc. rack units line up with the alternating hole pattern on the vertical rack flanges. One method is to look for the vertical holes with the smallest distance between them. These are the two holes only spaced 0.5" (12.7mm) apart. The bottom (or top) line of one 1U unit is right in the middle of these two holes (compare the below illustration). So this "thin bridge" is the line between Us. This 1U separator is sometimes marked on rack flanges with a horizontal printed or marked line (like a punched through small hole). When you start to screw in equipment, align the bottom (or top) of a first module with this line, or make sure that the top/bottom edge of equipment aligns with the middle between these two closely located holes so that mounting notches or mounting holes on front plates line up with vertical rack rail mounting holes. On racks with square rack holes, is is also common to mark the middle hole of a 1U three-hole group with a small side-notch.

Rack Depth and Cabinets

While the EIA has defined the front shape of a rack, the depth and what's behind the front rack rail is mostly left to the vendor. And as a rack's depth is the only variable in the equation, manufacturers are making their equipment deeper and deeper in order to max out what can be fit in a given rack slot. That's why 19" rack modules may be short enclosures or very deep. The actual usable depth in a rack is the "mounting depth", while the actual "channel depth" is usually a little deeper, close to the actual overall outer depth/ dimensions of a rack. Larger mounted enclosures, meaning with often added height and usually very deep, are commonly screwed to the front and rear posts, to properly support the heavy weight. When ball-bearing "Rack rails" are used as underpinning in this four-post bolting, the whole enclosure can be pulled out like a drawer for easier access. Such a setup is very common in high density data storage servers or disk arrays, where one enclosure may hold over 300 3.5" hard disk drives and max out the weight capacity of the form factor. So the effective rack mounting depth (in a 4-post rack) is the distance between the front of the front mounting flange to the back side/ rear side of the back mounting flange (or to the backmost unobstructed space within the mounting slot). Common mounting depth were 19" (for a square rack footprint), and 24". And while these smaller mounting depths are still common in traditionally designed rack modules, like measuring apparatuses or audio equipment, the bounding box of rack mount servers and computer systems has grown considerably in depth, to around 29" inches. With 2-post relay racks, the mounting depth usually only describes the "width" (when seen from the side) of the mounting post (the "upright"), but some racks feature more solid uprights, with about twice this depth. Now regarding the total outer envelope of a rack, the frame or cabinet, a complete rack is usually 42U or 48U (height units) tall, with half-height and other form-factors available for space constrained or small office applications. Depth and wide of closed rack cabinets can be much bigger than the actual footprint of the internal mounting rack, to offer room for cabling and attachments in front and on the side of the mounting area. For local IT or audio/IT equipment installations, it is common in back-office environments to have one half or full height rack, self-contained and sometimes sound-proofed, to be operated in lieu of a dedicated server-room.
Visual to-scale comparison between 10" inch rack and 19" rack dimensions
Visual to-scale comparison between 10" and 19" rack dimensions, CC-BY-SA Micropolis, adapted from cvdr, TomVocke et al.

Mounting flange shapes

Most racks use vertical upright mounting flanges in the shape of an "L", as having a "corner" on the rack, as the mounting flange, is a natural choice. Many flanges on open racks or in rack cabinets where it is possible to access the side of the rack cabinet, have rack holes on the front and on the side of the "L", allowing users to attach more stuff from the side or afix cable binders etc. Another shape found on rack uprights is a square "C" shape, where one "leg" of the "C", the front, is used as the front mounting area and the other sides of this three sided profile provide additional strength. The "C"-shape is related to the shape of the upright posts of a traditional 2-post rack. A third common profile, the "double rack strip", is a four-sided fold in the shape of an "ear" or "question mark" ("?") when seen from the top. It's a "C"-shape with an additonal leg on the rear part, forming kind of an "L"-shaped additional mounting front when seen from the side. This additional leg usually also has mounting holes, which can be used to attach, for example, fixed rails (L-brackets) or even rack equipment rotated by 90 degrees in case the rack opening to the side is of standard 19" (on some racks the depth allows 23" or 24" equipment to be installed from the side). One downside, of both, the "C"- and the "question mark"-shape, is that they quite often obsctruct something and prevent Rack rails or similar equipment to be installed easily or at all.

IEC Norm 60297 racks

The international (based in Switzerland) organization "International Electrotechnical Commission" (IEC, in French "Commission électrotechnique internationale)") has developed a nested series of norms to describe 19" rack mechanical structures and a system of 19" rack and subrack building blocks down to the PCB level - from rack cabinets (level IV) over rack equipment / module carrier chassis (level III) down to modules (level II) and PCBs (level I). IEC 60297 is largely equivalent or compatible to the standards described in EIA-ECA 310.
  • IEC 60297 defines the high level 19" rack "grid system", meant as an envelope or root document for the standards series
  • IEC 60297-3-100 (read 60297 Part 3-100; formerly IEC 60297-1 and IEC 60297-2) describes the basic dimensions of front panels, subracks, chassis, racks and cabinets (Level IV and level III)
  • IEC 60297-3-101 (read 60297 Part 3-101) describes Subracks and associated plug-in units, down to PCB sizes of the "Eurocard" format (level II and level I)
  • ... (omitted here, a number of detailing documents, regarding connector alignment, shielding, etc.)
  • IEC 60297-3-108 defines standards for "R-type" (ruggedized) subracks and plug-in units, i.e. ruggedized variants of the mechanical structures defined in 60297-3-100, but for heavy weight and large form factor, for large volume applications, e.g. cloud-computing servers, telecom servers
  • IEC 60297-3-107 similar to Part 3-108, but for Small form factor and light weight, for large volume applications, e.g. embedded systems, mobile/ubiquitous computing systems

Rack work safety

Although obvious, it might be helpful to point out that racks and rack cabinets are subject to the same laws of physics as any shelf or larger furniture. When heavy equipment is installed in a rack, users should try to place the heaviest equipment closer to the bottom. And while setting a rack up, users should install equipment from the bottom to the top, preventing racks from tipping over. Mobile racks on casters may feature leveling feet that should be extended once a rack has found a temporary final position, so that weight is taken of the casters and the rack establishes a more solid connection with the floor. In case multiple racks are placed side by side, check if baying tabs/ baying brackets can be used to join them together. Further, electrical equipment must be properly grounded - in racks potentially requiring running grounding jumpers to a grounding bus. And many electronic devices produce heat in near analog correlation to their power consumption. A 100W system may produce nearly 100W of thermal heat, equalling 341 BTU/h (British thermal unit/ per hour).

23-inch rack

The 23" rack is a wider rack format mostly adopted in the field of US telephony exchange infrastructure, sometimes referred to as the "Western Electric standard" due to mostly Western Electric using it. Hole spacing varies and only in some cases matches the 19" norm. A 23" rack is defined by its horizontal hole-to-hole spacing of 22-5/16" (or 22.312", equalling 566.7mm).

24-inch rack

The original EIA 320 standard actually defined three rack widths, 19", 24" and 30" with 19 inches being the preferred width. A 24" rack offers a rack opening of 22.750" and can be identified by its hole-to-hole spacing of 23.312 inches (23-5/16" or 592.1 mm).

2-Post Rack

also Telco rack (short for "telecommunications rack"). The 2-post rack is the "original rack". 4-post open racks or rack cabinets are essentially 2-post racks with extra supports or a protective enclosure.

3-Phase power

as of 2024, the use of 3-Phase power is on the rise in data-centers and server-rooms as it reduces the number of circuits required, optimizes power load balancing and allows a greater flow of power into individual racks.

4-Post Rack

Simpler racks are built with a frame of two vertical posts, the rack side rails. With heavier equipment, the front-only screwing mount might exert an excessive force to the front plate and a mounted enclose might bend or break. A 4-Post Rack solves this by offering a second pair of posts on the rear side of the rack enclosure as "Structural support". Bolting equipment to the front and the backside offers sufficient support for even the heaviest of equipment. Apart from additional mounting options, a 4-Post Rack is naturally also the base of most Rack Cabinets, speaking of a "4-Post Rack" with open constructions and an "Enclosed Rack" with side-paneled or closed constructions.

500 Series

the "500 series" is a form-factor specification for audio/signal-processing equipment and carrier Subracks, defined by Automated Processes, Inc. (API) during the 1970s. It is sometimes abbreviated as "API 500", stemming from an early EQ cassette sold and manufactured by API, the "API 550". The 500 series format was recently formally standardized and opened by API through their "VPR-Alliance". 500 series shelfes/ cabinets are usually named "housings" or a "lunchbox" (which is an API trade-mark), the latter emphasizing the fact that many users carry their setup of 500 series components (their "custom signal chain") with them on mobile recording situations or to music gigs. Each 500 series module (cassette) is 3U high and 1U wide and connects to a common power and i/o backplane installed on the rear of the rack. Many racks feature on-board PSUs, occupying 0.5 to three of the horizontal rack positions. With a full 19" enclosure offering space for 10 vertically inserted and horizontally organized cassettes (plus a little space left), an onboard PSU usually reduces empty slots to 8. Some 500 series cabinets allow cassettes to be inserted/installed horizontally, 3U high, so that when a "lunch box"-style enclosure is carried or placed vertically (19" perspective: "on the side"), with a handle on one side (then "the top"), the individual cassettes can be accessed and used in their intended upright orientation. API also offers modules for a "200 Series", featuring less tall 1:2 ratioed vertical modules and matching carrier chassis, with 12 slots (and a little) in a 2U 19" rack module.

ABC Analysis

Businesses or people handling hardware, installing racks or machinery sooner or later will speak of or hear the term "C-Parts". A "C-Part" is usually a part of low value, like screws in bulk or cage nuts. The term emerged from a business theory created by General Electric manager H. Ford Dickie in 1951. It's a classification system for arbitrary business objects and aims to help identify important goals and assets, differentiating important factors from elements of lesser or neglectable importance. ABC Analysis may be used to categorize business objectives, customers, costs or resources, as for example in inventory management and materials procurement. The labeling here with "A", "B" and "C" establishes as rank and is based on a given asset's importance in reaching set business goals, its value as such or for an intended outcome or its significance in a process.
  • A-Parts
    are usually of high value but ordered or procured in low volumes.
  • B-Parts
    lie in-between in terms of value and handled volume.
  • C-Parts
    mark the other end of the spectrum, are of low value but usually ordered or handled in high volumes.
Thinking from the end, a product, the category "A-Part" gathers the most valuable parts of a thing, bigger assemblies or structures. For example, in manufacturing a car, bigger automotive body parts could be "A-Parts". A-Parts make up for usually around 70-80% of the finished value of a thing. "B-Parts" form the intermediary class of parts of a product. In our example, it could be interior parts, cable assemblies or smaller body parts. In terms of part count, they make up around 30% of the total part number and contribute around 20% to the finished product value. "C-Parts" contribute usually below 10% to the finished product's value but cover over 50%, or up to 70% of total part count. C-Parts can be the headache of supply chain management: although they are of lowest importance, they need to be sourced, ordered, quality-controlled, handled, managed, etc. and after all are crucial to finish the product (may be "show stoppers"). In electronics, semiconductors, sometimes simple parts, carry the potential to halt a complete product line.

AC adapter

also known as "External Power Supply", "AC/DC adapter", "Wall charger", "Power adapter", "Power brick" or colloquially "Wall wart". An AC Adapter is a small kind of power supply and commonly the consumer or smaller version of a (more industrial, more enterprise-y) Power Supply Unit (PSU). AC adapters usually either come in the form of a corded rectangular box ("power brick") or as a case resembling a larger AC plug ("wall wart").
Different AC adapters (wall warts), EU and US plugs, and a brick-like laptop power supply
Different external power supplies as common for smaller consumer-grade devices. Tone-generators ("expanders", in music production), half-rack audio equipment and some shelf-based 10-inch rack components also often use these "wall warts".

Airflow & Cooling

While open racks in early telco installations might have gotten away without a dedicated airflow or cooling concept, in today's high performance world, in data centers and with scientific rack installations, a sophisticated design for proper airflow through and in-between components and an overarching cooling concept is paramount. With high density installations and high thermal power dissipation of individual components, moving heat away from devices and cool air or liquid into components is a challenge. On a per-rack level, cooling is either implemented as passive cololing or active cooling. In passiv colling, enclosures are either fully open, or feature specific intake/exhaust openings or have grooves, slotts or perforated doors and panels - allowing air to flow naturally, as warm air tends to go up. The upside of passive cooling is that it is basically maintenance free and produces no noise. The downside is that only a limited amount of air is moved, dissipating only smaller amounts of heat energy. Active cooling concepts, in contrary, use fans, blowers, air conditioning and ducts or even liquid, consume energy and must be maintained. The upside is, that cooling is more controllable and larger amounts of heat energy can be diverted while usually requiring less space and allowing equipment to be denser packed.

Many electronic devices, especially computer equipment, produce heat in near analog correlation to their power consumption. A 100W system may produce nearly 100W of thermal heat. Multiplying this Wattage by 3.412 gives 341 BTU/h. "BTU" (British thermal unit, from the imperial and US customary measurement system) is a common unit for the cooling power of air conditioning equipment.

On a data center level, one principle employed in many facilities is the concept of having hot and cold aisles ("Cold Aisle Containment"), meaning, for example, every second corridor in-between server racks is either a hot or a cold aisle, meaning a duct for either hot or cold air. Server racks in such setups are either partly open, without front and back door, or feature perforated doors with a specificed percentage of open area (usually over 70%). When air cooled servers or other equipment is installed in such racks, the device pulls in cool air from the front and blows out hot air on the rear. As the rack cages are the only link between the hot and cold aisles, every slot in a rack needs to be either occupied by a device or be closed with a blank filler plate, so that a rack, seen from one side, is a closed wall, separating the hot and cold areas. When unoccupied slots would remain open, the air could circulate through these apertures and the equipment would suck in warm air, rendering the colling concept uselesss. This is only one example of an airflow and cooling concept. Such designs may be based on the per-rack level, executed as a datacenter wide concept or even on a larger scale where a whole datacenteris embedded into a heat dissipation infrastructure, like building a site in cold world regions or where cooling water is found nearby.

As part of a hollistic management of a data center and here equipment cooling, airflow managament is one of many disciplines in operating an energy efficient facility. Computer room air conditioning (CRAC) units are precision AC systems built for data centers but based on the same compressor plus refrigereant layout that domestic AC systems are built on. Computer room air handlers (CRAH) have no compressor and use chilled water instead of a refrigerant. Further, it is important to manage and control how air is flowing in the datacenter and where it is led, in what amount and at which temperature. Over-provisioning air flow to the cold aisle can result in a cooling air bypass, meaning too much chilled air is made available. Under-provisioning cold air flow will lead to equipment running hot, building up hot spots in racks or datacenter areas, elevated fan speeds on equipment and potentially hot air recirculation. Specific structural elements (air skirts, air dams, curtains and even blank filler panels) can be used to guide and direct air towards equipment. Sensors can be used to measure intake temperature or detect hot spots. And an active monitoring in sophisticated systems may be able to regulate airflow and temperatures based on the current workload of running systems and their resulting heat dissipation on a per rack, per aisle or datacenter wide level.

One of the most effective technologies to cool IT equipment is liquid cooling. Systems may be either fully immersed into a cooling medium (Liquid immersion cooling) or ducts and pipes may direct a coolant (often distilled water) to heat exchanger blocks mounted on specific hot areas, like GPUs, CPUs, etc. (Water cooling). Once the liquid medium has absorbed the heat of a system, it it pumped or directed to a heat exchanger, like a radiator, to dissipate the accumulated heat and then recirculated back. Systems may be either active (using pumps) or passive (relying on hydraulics to circulate the coolant. In comparison with cold air cooling, using a liquid medium bears obvious challenges in operation, as electric equipment and usually ductible liquids are natural opponents. But while the elevated complexity of liquid cooling or the potential of damages or leaks are downsides, the high efficiency of the technology is the reason for more and more datacenters switching to this technology - either in smaller specific areas or larger parts of the IT stack.


or "Aluminium" in British English and many European languages, is a chemical element denominated by "Al" from the group of metals, often abbreviated as "alu", a light metal with a visual resemblance to silver. As aluminum is not containing iron, it is grouped in metallurgy as a non-ferrous metal. In comparison with precious metals like gold or silver, aluminum is a base metal and is quick to react on clean surfaces with air and water to build an aluminumoxide layer. This means that aluminum has to be properly primed before a number of coatings would build a lasting bond with the metal's surface. Aluminum is an abundant material and thus popular in construction, computer cases, electronic boxes or IT racks. As aluminum is more expensive than steel or iron, aluminum is usually used only where wheight is a defining factor, for example in aerospace applications. In addition, the cost factor is influenced by the fact that aluminum is softer than steel and requires thicker or more solidly designed structures to achieve similar rigidity. While it is usual to refer to all these embodiments as "aluminum", the actually used metal is mostly an alloy where aluminum is the primary element. Common aluminum alloys are aluminium–magnesium alloys (AlMg), like "AlMg1", "AlMg3", or alloys with a small amount of manganese ("AlMgMn", like "AlMg4,5Mn0,7"), aluminium–magnesium–copper alloys ("AlMgCu") or aluminium–magnesium–silicon alloys ("AlMgSi"). Apart from aluminum alloy sheet metal blanks, formed and treated in a number of ways, "aluminum" is often extruded into profiles, for example the popular T-Slot alu beams.

Aluminum priming

Aluminum has many desirable properties but also the undesirable property of building an aluminum oxide layer on the very surface, by itself and quite quickly. And while this natural process is good in protecting the aluminum against environmental influences, it is also the reason why aluminum is relatively difficult to paint or coat. Enamel and acryllic varnishes will simply not adhere ("stick") to untreated aluminum and paint is destined to come off in places. Mechanical aluminum preprocessing, like sanding or washing, even just before painting, won't perfectly remove the oxidized layer. Thus, in painting or coating aluminum, a specialized primer solution has to be applied, sprayed or painted, in order to allow the varnish to build a firm bond with the metal.

Angled Rack Stand

Especially in audio and music production, the use of smaller, sub-rack-type angled cabinets is common, a type of rack that features an angled front, inclined towards the user. These racks are mostly used to rack only a few devices, around 3U to 6U. Designs of angled racks are either simple metal frame layouts in the form of open racks, or closed small cabinet styles, often wooden racks, where the upper part is slightly slanted. Open rack types and cabinets usually feature an inclination of around 20-40 degrees backwards (negative, away from the user) so that the front-plate of installed equipment is easily reachable and readable by the user. One common problem with the slanted design is that the bottom rack slot is not able to fit a full-size (deeper) rack device, as the slant means the depth of the slot is limited by the bottom side or base of the rack stand / cabinet. Some designs thus feature a small elevation of the bottom-most slot to add more depth to this lowest rack position. Small card cage-style enclosures featuring a hinged handle are usually called desktop (instrument) enclosures. Compare Desktop Rack.
Three angled rack designs: an open-frame angled rack stand, a wooden slanted audio rack cabinet and a hinged handle desktop instrument card cage.
Three angled rack designs: an open-frame angled metal rack stand, a wooden slant cabinet audio rack and a desktop hinged handle instrument card cage that can optionally be carried around ("mobile enclosure").


Anodizing is a surface finishing process for metals. Anodizing is one form of surface passivation and part of a number of processes of intended anodic surface oxidation. Different techniques of surface anodization exist, with an electrolytic passivation process sometimes known as "Eloxadizing" (a German word creation) being the most popular. The Eloxal anodizing treatment actually transforms the uppermost surface of an object by growing an oxide layer, usually to the same amount "into" the surface as it "grows" to the outside. This actual transformation of the surface is different from galvanic processes where a different material (usually zinc) is laid (applied) onto the metal surface. Anodizing is a measure to prevent rusting (corrosion). It is possible (and popular) to create colored anodized aluminium.

ATA road case rack

The abbreviation "ATA" is short for "Air Transport Association of America" (ATA) the older name of today's "Airlines for America" (A4A), a central organisation for lobbying and standards in the aviation industry. One standard brough forth by ATA was the "Specification 300" dating back to 1960, describing a case that is "Reusable for a minimum of 100 round-trips usually fabricated out of plastic and / or metal" with "(a)ll hardware, including fasteners used to secure a lid closed, shall be recessed, flush or guarded so that no protrusions could cause damage to the container or to other goods shipped in the same conveyance" (Specification 300, 2008, pages 31 and 22). For touring bands and troupes, such cases became the go-to boxing for equipment, and also for readily installed 19" rack based equipment. This is where the term "flight case" came from, or from being on tour, "road case". The common appearance in black wood or plastic with silver aluminum or steel corners became synonymous for a "flight case", even when it doesn't adhere to original ATA specs. For easier handling, ATA compliant rack cases are usually not 42U / full-height tall, but shorter, of around 21U and on caster wheels. Inside, flight case racks are often insulated and shock-proofed, via damping material or spring mechanisms. Cabling from one of those racks to another or to amplifier, speaker or light equipment is usually done via Harting connectors and thick "cable snakes" ("cable assemblies").

Automatic Transfer Switch

sometimes "Auto Transfer Switch", is a feature of Power Distribution Units (PDU) to offer power redundancy on the PDU level. Having a PDU connected to multiple power sources, "ATS" is an in-built mechanism of the Unit to switch from one power source to the other in case of failure of one.

Baying brackets

Rack frames and enclosures usually offer a number of different options to install, bolt down or afix a system to a wall, the floor or movable bases - via a number of optional mounting points. Many mounting points are slotted apertures, allowing screws to be put through, often in a bayonet mount -style fashion. Some racks feature "baying brackets", "baying tabs" or other systems ("joining kit") to connect racks side-by-side or over-another. Joining multiple racks improves overall stability, preventing racks from tipping over or move. Especially in A/V and mobile/roadshow uses, it is common to have smaller racks, half-height 21U or similar, for practical reasons, that are then stacked two racks high and bolted together for temporary installations.

Bend Radius

is a term from the field of sheet metal forming. Similar to folding paper, sheet metal can be folded or "bend" to form flanges into angles of 90 degrees or arbitrary acute or obtuse angles. The bent area, a linear ridge, has an inner and an outer radius - depending on which side of the resulting flange one measures. And while designers tend to think that it is possible to form, for example, a box from sheet metal without any bending radii, the reality is that sheet metal bends always form a rounded corner, not a perfectly sharp edge. When a flat sheet of metal is formed, for example in a press brake, the punch is only able to press the metal into the die to a certain degree, otherwise the metal may crack, either already during manufacturing or later, down the line, when some form of stress is applied. That's why a minimum bend radius is used for all bends during manufacturing. The actual bending radius depends on the thickness and the malleability of the formed material. For steel, the rule of thumb is to use bend radii equal to the material thickness, meaning 1.5mm steel asks designers to incorporate inner bend radii of 1.5mm into their designs. Aluminum is softer and common bend radii for this material are either a factor of 0.8 or even down to half the material's sheet thickness. Similar is true for copper or brass.

Bend Relief

is a term from the field of sheet metal forming. When sheet metal is bent, for example in a press brake, the area where the metal surface transitions from a flat area into the bend, or in-between two bends, depending on material properties, the material may tear at the bend's side. In order to prevent tearing, the general rule in sheet metal bending is to apply a "bend relief", "relief groove", "relief notch" or "relief gap", a small area around a bent corner that is widened to allow the material to distribute the stress of the bend more evenly. Bend reliefs may be "straight" rectangular notches with right-angled corners, "rounded" slots ("obround") or circular ("round") cutouts covering the high stress area. The width of a relief is usually at least half the thickness of the used sheet material and extend a fraction behind the bend. After fabrication, bend reliefs also contribute to the longevity of a workpiece, as proper reliefs help prevent fracture propagation, improve durability under vibration and surface coatings are spread more evenly as fabrication without bend reliefs may produce burrs or sharp points. On metal enclosures, in corners where three flat surfaces meet, the bend recess or corner relief can often be noticed as a small hole or eye.
Inner bend radius, bend relief gab and corner relief cutout in sheet metal forming
Sheet metal forming: the inner bend radius, bend relief gap and corner relief cutout with flanges

Blanking Panel

Many rack mounts or cabinets offer more space than is occupied by front-facing equipment. In order to tidy up the appearance of a rack, optimize Airflow or secure critical cabling, an unspecific simple piece of sheet metal or extruded aluminum can be bolted into free units of the front-, rear-, or both sides of a rack. Blanking Panels (also "blanking plate" or "Blank Filler Plate") are sometimes customized with DIY modifications (drilled holes, sawn openings, etc.) to serve specific mounting needs of non-standard sized devices. Some blank panels are metal and screwed in place, while others are plastic or other flexible material, that is attached in many height units at once, like a rolling shutter on a window, and fixed to the rack rails with a simple clip-on system. Some blank rack panels may play an integral part in the structural struength of a rack system, while with other racks, using a blank filler panel is optional, a matter of preference, visual style or a needed element in an airflow concept.

Brush Panel

Blanking plates come in many forms, with ventilation openings, with cable ducts, cable management hooks etc. A "brush panel" has a large opening with an inserted horizontal brush curtain, so that cables can be easily run through it and the brush bristles acting as a simple seal around the cabling, helping to control air flow inside a cabinet.


abbreviation for "Baseboard Management Controller" (sometimes "Board Management Controller") is a small dedicated computer-system (a microcontroller) embedded into the mainboard (baseboard) of systems that are usually operated remotely, like racked server systems in datacenters or NAS systems in off-site locations. A BMC is the central hub for sensor data, system state, power cycling and other means to supervise and control the host system. A BMC can be part of the IPMI control stack and may communicate over IPMI protocols. Often BMC networking is sperate from host system networking, adding an additional layer of security by using a dedicated communication network or serial connection. The software (client) counterpart of the BMC is the "BMC Management Utility" (BMU), usually a command-line tool to communicate with a remote BMC. Some BMCs offer a web interface.

What are C-Parts?

The term C-Parts emerged as part of a structured approach in business theory called "ABC Analysis". In practice, ABC analysis found application in material management and supply chains. C-Parts according to the ABC classification are parts of low value but parts that are usually handled, ordered or procured in high volume, usually in bulk, are difficult to order and manage and often impossible to substitute - despite their low value. Small things like washers, nuts, screws, bolts or cage nuts are usually labelled as "C-Parts". In electronics, simple semiconductors are dreaded C-Parts and may bring whole product lines to a halt.

What is a cable?

A cable is a collection of wires. While the terms cable and wire are often used interchangeably, there's actually a difference. The wire is the conductive metal part, often copper, aluminum or an alloy, which may be made of a solid extruded piece of metal or multiple strands. A wire doesn't necessarily have to be protected on the outside by a non-conductive material. However, a cable is a collection of wires where each wire is protected (insulated) against contact with the other wires in this cable by some form of non-conductive material, individual thermoplastic sheaths around each wire or by embedding all individual wires in some common insulator material. The collection of wires inside a cable is usually wrapped together by some form of outer cover, made from the same or a different insulator material as used for the internal wires. It is common to have wires and/or cables being wrapped by silicone, rubber, textile meshes, thermoplastics, vinyl, rubber, polyurethane or heat-shrink plastic. With many cables, complex cabling or in combination with movement, cable and wiring systems may need some form of "Cable management".

Cable Arm

Some rack cabinets or rack mount equipment features specially designed "cable arms" as part of cable management. When installed equipment is, for example, placed on rack rails and is later pulled out for maintenance, the multitude of cables routed to a server or storage enclosure might get in the way, get damaged or get pulled out. In order to prevent cables getting stuck or in the way, all cables during setup of a system are placed on specific articulated cable arms, a system of links and joints, bolted to the enclosure or the rack cabinet. When a system is pulled out, the cable arms swings out accordingly, supports the cable and manages them in a predefined way, so a system may continue running uninterrupted or the cabling loom of the installation remains intact.

Cable Classes

Computer networking Patch Cables are divided into several classes, where each class defines the amount of shielding employed and the maximum frequency rating. Electrical ("copper") cables are categorized as:
  • Class A: Up to 100 kHz using Category 1 cable and connectors
  • Class B: Up to 1 MHz using Category 2 cable and connectors
  • Class C: Up to 16 MHz using Category 3 cable and connectors
  • Class D: Up to 100 MHz using Category 5e cable and connectors
  • Class E: Up to 250 MHz using Category 6 cable and connectors
  • Class EA: Up to 500 MHz using Category 6A cable and connectors (Amendments 1 and 2 to ISO/IEC 11801, 2nd Ed.)
  • Class F: Up to 600 MHz using Category 7 cable and connectors
  • Class FA: Up to 1000 MHz using Category 7A cable and connectors (Amendments 1 and 2 to ISO/IEC 11801, 2nd Ed.)
  • Class BCT-B: Up to 1000 MHz using with coaxial cabling for BCT applications. (ISO/IEC 11801-1, Edition 1.0 2017-11)
  • Class I: Up to 2000 MHz using Category 8.1 cable and connectors (ISO/IEC 11801-1, Edition 1.0 2017-11)
  • Class II: Up to 2000 MHz using Category 8.2 cable and connectors (ISO/IEC 11801-1, Edition 1.0 2017-11)
Fiber-Optical cabling is defined in a number of OM ("Optical Multimode") classes defined by respective ISO, TIA/EIA and IEC norms:
  • OM1: Multimode, 62.5 μm core; minimum modal bandwidth of 200 MHz·km at 850 nm (retired)
  • OM2: Multimode, 50 μm core; minimum modal bandwidth of 500 MHz·km at 850 nm (retired)
  • OM3: Multimode, 50 μm core; minimum modal bandwidth of 2000 MHz·km at 850 nm
  • OM4: Multimode, 50 μm core; minimum modal bandwidth of 4700 MHz·km at 850 nm
  • OM5: Multimode, 50 μm core; minimum modal bandwidth of 4700 MHz·km at 850 nm and 2470 MHz·km at 953 nm
  • OS1: Single-mode, maximum attenuation 1 dB/km at 1310 and 1550 nm (retired)
  • OS1a: Single-mode, maximum attenuation 1 dB/km at 1310, 1383, and 1550 nm
  • OS2: Single-mode, maximum attenuation 0.4 dB/km at 1310, 1383, and 1550 nm

What is a Cable harness?

Often when cables are bound or assembled together as part of a thicker bundle of cables, some sort of "Cable management" is used to improve maintainability, operational safety and reduce faults, by preventing wiring failures or people or equipment getting entangled in cables. Some simple form of cable management is pulling individual cables through cable ducts, but during manufacturing, when the same wiring setup is repeated many times, this may lead to wiring faults or unnecessary stress on individual cables. Thus, in the early 20th century, it became popular to bundle cables in ad-hoc bound or pre-assembled cable harnesses. Sometimes different terms are used to describe different types of Cable harnesses ("wire harness", "wiring harness", "cable assembly", "wiring assembly" or "wiring loom").
A "wire harness" is usually the most simple form of a cable assembly. Here, a bundle of usually single core wires (so its probably better to speak of a "wire harness") is assembled into one bundle of wires, each isolated against each other but bundled. Over the years different techniques of cable lacing were used, with rope, with tape, with cable ties, knitting in elaborate patterns. A wire harness usually has each single wire visible.
More elaborate are "Cable assemblies", where a number of cables are combined into one (thicker) sleeve. In contrary to a wire harness, a Cable assembly usually is a bundle of cables, meaning the individual "wires" are actually multicore cables. A cable assembly has all included cables protected by a single outer protective sleeve, giving the appearance of one thicker cable.
A "Cable loom" (sometimes less exact "wiring loom", or "electrical loom") in turn is a bundle of multiple Cable assemblies, wire and combinations thereof into one complex system. Many cable looms have the appearance of wiring harnesses, but each component is a multicore cable or wire connection.
Wire harnesses and Cable looms often are pre-fabricated on dedicated pinboards, where a defined routing layout is used to route wires and cables into a specific pattern, giving the resulting assembly a specific shape, where individual wires or cables may have different length, following defined paths or breaking out from the harness/loom at specified points. After a build, technicians can tests such pre-assembled harnesses/looms by connecting the assembly to test circuits. During installation of such assemblies, having a tailored bundle decreases installation time, prevents mistakes and helps with the standardization of work processes.

Cable Management

Cable management as a general practice is the process or guideline of laying, routing and pulling cables in a tidy and organized way. With moving systems, where cabling is applied outside the moving structure, like with robots or when connected modules move relatively to each other, carefully controlling cable flex improves security and maximizes cable life. Thus, cable carriers for cables and hoses are commonly used on industrial robots or in rack equipment to optimize cable management. From a more general perspective, cable management is an important aspect of mechanical and electrical engineering, inside machines and housings and cabinets. One important technique is to gather a number of cables in "Cable ducts" or assemble multiple cables as part of a "Cable Harness".

Cable Raceway

also "cable rack". A light steel frame structure, usually located over racks or rack cabinets to lead cables to individual racks, commonly suspended on threaded rods from the ceiling. In data centers, usually network cabling is done separately from power cabling, so that high power connections don't induce a charge into network cables. It is common to place low power cables, like networking cables, in overhead cable ducts, cable raceways, and have mains power be connected from under the elevated floor.

Cabling Standard

There are a number of standards regarding data center cabling, giving advice and setting best-practices for (structured) rack cabling, rack to rack cabling, row cabling, and data-center wide / backbone or facility cabling.
  • ISO/IEC 24764
    more specifically "ISO/IEC 24764:2010(E)" specified generic cabling systems for data centers, supporting a wide range of communications services. It covered balanced cabling and optical fibre cabling. It was based upon and referenced the requirements of ISO/IEC 11801 and contained additional requirements that are appropriate to data centres in which the maximum distance over which communications services are distributed does not exceed 2000 meters. This now replaced standard has been revised to form "ISO/IEC 11801-5:2017", incorporating corrigenda and amendments of "ISO/IEC 24764:2010/Amd 1:2014".
  • ANSI/TIA-942
    Telecommunications Industry Association (TIA) ANSI/TIA-942-B (in revision B), officially the "Telecommunications Infrastructure Standard for Data Centers" is an "American National Standard" (ANS) setting minimum requirements for data center infrastructure. It certifies that design documents of compliant equipment have been reviewed for conformity to the design criteria, aligned with a certain "rating level". For facilities, that design documents have been physically onsite inspected and been verified and practices like modularity have been implemented according to the standard an a respective "rating level". These "rating levels" refer to the standard's four rating levels: "Rated-1: Basic Site Infrastructure2, "Rated-2: Redundant Capacity Component Site Infrastructure", "Rated-3: Concurrently Maintainable Site Infrastructure" and "Rated-4: Fault Tolerant Site Infrastructure", the highest level and most redundant type of rating. As of 2024, the standard is again under revision and about to be released as "ANSI/TIA-942-C" (revision C).
  • ANSI/BICSI 002-2014
    is an "American National Standard" (ANS) for Data Center Design and Implementation Best Practices. It is usually regarded as complementing the ANSI/TIA standards, as BICSI publications tend to be more detail-oriented than TIA standards. BICSI as such provides detailed guidelines for the layout of mechanical, electrical or thermal systems of a data center, and taking related topics like security considerations into account.

Cage Nut

A "cage nut" ("captive nut" or "clip nut") is a square nut wrapped in a small metal sheet (a "cage"), having two or four small spring clips which grip to square holes of vertical rack rails. As defined in various standards, 19" racks may feature threaded or unthreaded holes, but threaded holes may get damaged and unthreaded holes require a nut and a second tool to hold the nut while fastening the screw. For quick installation or tool-less mounting, square rack holes were introduced and with these a "cage nut" is needed to mount equipment. Once installed, the cage nut works like an adapter and can accept either M6, 10-32 or 12-24 screws - whatever fits the installed equipment. The cage of such a nut sits a little loose inside the square rack hole but locks tight once the equipment is screwed in. Note that such cage nuts, due to how deep the little metal collar is, need to fit the thickness of the rack strip they are clipped into. Rack strips may be thin 1mm metal or thick heavy duty material of 2mm or more.

The thing with cage nuts is that they are manually quite difficult to insert into the full square holes of a rack strip. The little spring collar may be hard to compress, can snap back and injure the rack worker. Or cage nuts are quick to pop out while inserting them, fall down and make all sorts of headaches. That's why many companies have tried to improve the situation. Some rack equipment, like heavy duty server rails, come with built-in rack hole fasteners that are easy ("tool less") to insert and lock. And for installations based on loose cage nuts, a number of cage nut tools and helpers are offered. Some have tried to improve the design of cage nuts themeselves. Former times Micropolis neighboring company Chatsworth Products offers the patented "clik-nut", where an elongated lever arm helps to compress the cage nut cage. Others have reversed the nut into a threaded rod that is inserted from the back of the rail and makes the rod stick out to the front, offering guidance during installation and subsequently easy fastening with a thumb-size nut from the front of installed equipment. These ideas and innovations come in a number of shapes and forms and it's up to the user to decide which one is best.

Cantilever Rack or Shelf

Cantilever is a synonym for a simple beam or statically loadable arm holding a weight. In logistics and storage, Cantilever Racks are open racks with shelving levels made up of horizontal steel beams, usually used to store long goods, like carpet coils, wood packages, sheet metal. In a datacenter context, a cantilever rack or shelf is a type of Rack Shelf with a slanted support structure or side beam, and as any rack shelf can be used to store arbitrary smaller hardware and/or devices. Sometimes an interlock system (grooves or cut-outs on the shelf bottom) is offered to attach smaller sub-assemblies to hold (consumer) devices, for example large numbers of mobile phones, side by side in high density, for Ci/CD ("continuous integration and continuous delivery") applications, where apps or software is tested on many hardware platforms to guarantee compatibility, etc. Cantilever Racks are sometimes customized with DIY modifications to serve specific support needs.


Especially smaller racks or subracks are often put on casters. Casters are either one way or rotatable, can be heavy duty or limited in weight capacity. For heavy duty equipment, casters are often complemented with rectractable leveling feet, short threaded rods, that can be extended further than the casters' height, lifting the rack from the casters / ground, making it immobile but adding much to a firm stand of equipment, even allowing the user to compensate for smaller floor uneavenness.

Cold-rolled steel

also known as "Cold-formed steel" (CFS) or "cold worked steel" is any formed steel that is shaped in a processes carried out below recrystallization temperature (of steel). In raw material, like sheet metal, cold-rolled (in difference to "hot-rolled steel") usually describes how the metal was brought into the flat raw (sheet) shape. Shaping or "forming" can mean rolling to reduce thickness (as noted before), but also subsequent processes, like pressing in a press brake, punch press or stamping press, bending (folding) in a brake press or Cornice brake et cetera. Cold-forming usually leads to heightened strength and toughness but as steel in its various mixtures and alloys is graded in many aspects, the more colloquial terminology "cold-rolled steel" is often used as a marketing term to emphasize the robustness or strength of the used material in rack equipment. Cold rolled is one of many features to describe steel types and grades.

Co-Location Cabinet

depending on style often written as "CoLocation Cabinet" in CamelCase, is a type of 19" (computer) rack that features multiple, often three, front doors, offering three isolated racking compartments, with heightened security features, locks and an overall sturdy design, so that data centers offering access to equipment by their tennants can allow access to the data-center floor and a customer's rack without exposing other tennants' equipment to unauthorized access.

Commercial off-the-shelf

abbreviated as "COTS", sometimes varied as "commercially available off-the-shelf" is a term to describe products that are ready-made, don't need preparation or adaption prior to deployment and are usually easily available with short lead times. COTS is the opposite of tailor-made, custom-made, or bespoke solutions. Using a Blanking Panel, drilling and sawing an opening to it to build a DIY router mount would be using an "off-the-shelf" blanking panel to build a custom rack-mounting kit. One well known abbreviation related to COTS is "Mil-COTS", refering to COTS products for use by/ or used by the U.S. military.

Console Drawer

Slide out combination of computer display, keyboard and trackball or mouse, usually designed in the form of a 1U or 2U rack mountable drawer. Most comnsole drawers are single screen but some designs offer multiple displays which slide out horizontally. Datacenter technicians can then stand in front of a rack and operate a console terminal without carrying a separate laptop or other terminal. A console drawer is an on-site local terminal solution (cmp. "Console Server" or "remote management").

Console Server

also known as "console access server", "console management server", "serial concentrator", "serial console server" or "terminal server" is a device that connects to the RS232 serial port of networking (router, PBX), computer (servers) and enclosure management devices (PDUs, Rack management card) and tunnels their serial communication via Network (LAN, TCP/IP, or POTS/ "Plain Old Telephone System") to a remote management terminal ("Serial over IP"). While the more general term is "Terminal Server", in datacenter lingo the more common term is "Console Server" as the terminal server is commonly used only to connect to console ports (COM ports) of connected hosts and tunnel system DevOps sessions via Telnet or SSH ("remote console").

Cornice brake

Simple Cornice brakes are hand operated bending benches for metal forming. A sheet metal blank is placed on a flat bed and affixed by some clamping mechanism, holding it firmly. One part of the formerly flat bench is hinged and can be rotated around so that one flank of the sheet metal is bent. Some Cornice brakes operate circular while others have a more ellipsoid hinging mechanism, compensating for the deformation of the metal during the bend. Modern large computer controlled panel bending machines ("panel bender") work similar, but combine the flat feeding mechanism with a computer controlled bending stamp. The difference to a press brake is that a "Cornice brake"-like mechanism doesn't press the metal sheet into a tool but produces the bend by bending the metal against an edge to produce a radius ("swivel bending"). Modern machines are also often embodied as "combination machines", adding the ability to laser cut inserted workpieces.


short for "Computerized Numerical Control", a "numeric control" unit that is improved with or embodied as a computer system. While the term originally meant the control unit only, it is usually a "pars pro toto" term for a whole machine or a machine process, often meaning milling on a multi axis lathe or milling machine. For a more detailed explanation of the term and its origin refer to NC.

DCIM (Data Center Infrastructure Management)

Usually software solutions that allow data center operators to log, monitor and control d data centers infrastructure, usually down to the row and rack level. Monitoring security, environmental conditions and metrics like power consumptions allows operators to employ best practices and enforce regulations, for example, for power consumption or energy consumed for climate control.


Deburring is a term most commonly found in metal works. When sheet metal is cut, edges tend to be very sharp, with near perfect 90 degree angles and no roundig or chamfering. Especially when metal is cut using water jet or laser, edges are sharp - often with parts of the material molten or vaporized within the kerf setting on the edge, forming sharp burrs. The process of deburring removes these burrs and sharp edges, using a number of different processes which depend on the characteristics of the work piece. When a work piece is intended to be painted after cutting, deburring is important so that the paint is able to form a uniform layer of coat, also on edges. Sharp edges tend to deteriorate the quality of coatings. Deburring is usually not done on plated metal (like galvanized, zinc plated metal) though, as the abrasive process of mechanical deburring would remove the protective layer.

Desktop (Rack) Enclosure

Many applications require 19" racks or subracks to offer a "desktop mode", where the case can be easily placed on a desk and installed racked devices or installed cards can be easily operated, knobs reached and tweaked, etc. Common designs for desktop racks meant for installation of 19" rack modules are smaller angled stands (or slant cabinets), where the front of the cabinet is facing the user at a fixed angle for ergonomic operation. Compare Angled Rack Stand. Card holders ("card cages") are usually even smaller when designed for dektop use and very often feature a hinged handle that can be used to carry the case around and then tip-up the enclosure to face the user when placed on a desktop. An illustration of such a mobile card cage is in Angled Rack Stand. These card cages usually have a traditional instrument design style, and while 500 Series and similar audio gear usually features the same vertical 19" rack compatible installation layout (borrowed from normed card cages), desktop audio enclosures often follow a "lunchbox design", with a handle on one side.

DIN Rail

is a horizontal metal mounting rail usually used for mounting PLCs (industrial control equipment), relais, motor controlers and circuit breakers. It emerged in Germany, where it was first implemented during the 1920s and later elaborated into an industry standard (German Norm) by "Deutsches Institut für Normung" (DIN). As of 2001, DIN EN 60715 defines four variants, which have since been adopted as international (IEC) standard (IEC/EN 60715). The most well known DIN Rail is the top-hat shaped "Top hat rail" ("Hutschiene"), with 35mm wide types available in two depth dimensions and one smaller 15mm wide variant. The 35mm wide variant is known as "TS35 rail" in the USA. The fourth standardized type is a "G"-shaped rail type. A "C"-section rail type sometimes used is not part of the standard. Similarly to 19" rack mounting equipment, modules on a top hat rail are measured in "module widths", usually abbreviated as "M" (or less often "standard units", "SU").


In order to protect metals against corrosion, deep drawn metal parts or whole assemblies can be coated by "dipping" (submerging) the whole structure into a bath of liquid coating solution. When the structure is lifted out of the bath again, a thin layer of the coating material dissolved in the liquid has settled on the surface of the object. In electrophoretic deposition (EPD), part of an overarching category of coating processes known as "chemical solution deposition" (CSD), an electrostatic charge is used to optimize the adhesion of the particles dissolved in the coating liquid - similarly as electrostatic charging is used in powder coating. Dip-coating is very effective to guarantee that a protective film has built on every surface and in every crevice of a structure. The quality and layer-thickness of the applied coat depends on environmental factors like temperature, air pressure and humidity or the motion of the structure inside the bath. In automobile part production, it is common that a whole car body is drawn and rotated through large basins to ensure that the anti-corrosion liquid has reached every little unevenness of the part, such as with small trapped details or on weld seams or weld points. Once the dip coat is applied, the coat is cured in a heat treatment, baking the coat onto the processed surface (stove enamelling). Dip-coating may be part of a multi-step anti-corrosion treatment, where the Dip-coat layer acts as a first ceramic-like protective layer ("dipcoat priming") which is then powder coated in a second step to perfect the surface protection.


short for "Earth Bonding Point", a connector used by technicians to establish a ground connection to prevent an electrostatic discharge. See ESD.


Short for Electronic Industries Alliance (EIA). Rack equipment is standardized in document "EIA/ECA-310" for "Cabinets, Racks, Panels, and Associated Equipment". As the EIA ceased operations in 2011, ECA continued her work, and in turn ECA is expected to merge with the National Electronic Distributors Association (NEDA) to form the Electronic Components Industry Association (ECIA).

Electromagnetic compatibility (EMC)

electrical equipment usually produces unwanted electromagnetic energy and EMC labels techniques to mitigate or control the amount of emitted electromagnetic energy or to allow equipment to work inside its defined parameters while being immersed in a specific electromagnetic environment, dimming effects such as "electromagnetic interference" (EMI).

Electromagnetic interference (EMI)

in radio systems or equipment sensible to radio waves, electromagnetic inference (sometimes "radio-frequency interference (RFI)") labels unwanted induction of interference (disturbance) into a system, computer or, more general, any electrical circuit. Various strategies can be employed to mitigate EMI, for example, electrical ducts can be laid out in way so that they are less prone to act like an antenna. In EMI shielding (EMC protection, hardening), the enclosure of equipment is an important element, with choice of material, design, coating of surfaces and sealing playing an integral part.

Enclosed Rack

A type of closed rack cabinet, opposite of an open rack frame, used to better protect mounted devices.


abbreviation of "Electrostatic Discharge" are short and often intense electrical breakdowns between two differently charged surfaces or objects (potential differences). This is what happens in thunderstorms, where lightning is the electrical breakdown through the atmosphere between the charged cloud region and earth's ground. While lightning in thunderstorms is very intense, in this case even surpassing the breakdown voltage of the atmosphere, electrostatic discharge on a smaller scale can be similarly dangerous. Working with computers, microelectronics or more general with ESDS (Electrostatic Discharge Sensitive Device, "susceptible to electrostatic discharge") preventing an electrostatic discharge is commonly tried to be mitigated as the unwanted flow of high current can or will damage equipment, can lead to costly losses, fires or even fatal injuries. A common source of electrostatic charges is friction or movement of two differentially charged surfaces on or against each other ("Triboelectric effect"). In electronics, ESD aware workplaces ("Electrostatic Protected Area") thus are equipped with special flooring material, chairs, desks and equipment (cords) to ground people and tools and people working with ESDS may be required to wear ESD aware clothing and shoes. Some equipment racks feature dedicated and standardized "E.S.D. Bonding Points" (also "EBP", "Earth Bonding Point") where service technicians working on equipment can make proper conductive ground contact via a snap fastener (button, popper) prior to touching equipment within a rack. Areas where common ground can be connected are sometimes marked with a symbol, the "ESD Awareness Common Point Ground" symbol. With doors or other disconnected metallic elements of a rack cabinet, vendors need to make sure that a whole enclosure is grounded, for example by installing grounding jumper cabling. Installed equipment is grounded by having it rest on bare, not insulated conductive metal and/or offering a groundign bar (grounding busbar, grounding point) for equipment to be grounded via a dedicated wire. For datacenter infrastructure, ANSI/TIA standard 942 includes grounding rules, organized into four rating levels, with level 1 not requiring a dedicated facility wide grounding infrastructure and level 4 being the highest rating ("Fault Tolerant Site Infrastructure").

There are various guides and norms in relation to ESD. ANSI/ESD S20.20 and related norms are the American National Standards Institute's standards for the development of an electrostatic discharge control program, covering electrical and electronic parts, assemblies and equipment. Similarly DIN norm DIN EN 61340 and its related regulations are ESD standardization bodies within the EU region. There are various theoretic models to describe ESD issues. In general, there's a differentiation between "conductive" and "dissipative" surfaces. Conductive surfaces exhibit resistance lower than 10KΩ, R < 10KΩ. Dissipative or electrostatic dissipative surfaces exhibit a resistance between 10KΩ and 100GΩ, R ≥ 10 KΩ < 100GΩ. Insulators and insulating material and surfaces exhibit a very high resistance, usually above 1011Ohms. While raw metal surfaces are usually part of the first category of conductive materials and surfaces, coated or treated surfaces are often less conductive and thus are rated as being either dissipative or insulated. In powder coating metal, for example, a thin layer of plastic (epoxides or polyesters) is laid on a surface and thus these coatings act as an effective insulator. For ESD aware applications or for use in "Electrostatic Protected Areas", surface coating thus may be done with special electrostatic dissipative powder coating material containing a conductive additive.

Industry Electrostatic discharge (ESD) symbols
ESD symbols, indicating susceptibility to and protection against electrostatic discharge (ESD) and the symbol for a common grounding point


Short for "Electrostatic Discharge Sensitive Device", see "ESD" for more.

ETSI rack

The ETSI rack (sometimes "21-inch rack") is a standard defined by the European Telecommunications Standards Institute (ETS, in document "ETS 300 119") for rack mounting infrastructure. One difference to common EIA-310 rack is that the ETSI rack is slightly wider, with front-plates measuring 21.1in (535.00mm) wide and a rack opening of 19.68 inches (500.00mm).
ETSI rack vs. EIA-310 rack dimensions
ETSI rack vs. EIA-310 rack dimensions, CC-BY-SA, by Jahoe

ETSI bracket

An ETSI bracket is an adapter used to install EIA-310 (19-inch rack equipment), which is about 2 inches less wide, into the larger rack opening (500mm) of an ETSI standards based rack frame or enclosure. Some ETSI brackets are embodied as "rack ears", screwable to the side of 19" equipment, while other ETSI adapters are build as frames being slid over the smaller 19" devices.

Extruded Aluminum

The process of extrusion (from Latin "extrudere", to expel or drive out) is a mechanical process for metal forming and also plastic forming. It is used to create profiles of a given shape (or cross-section) by pressing a raw block of material through a die (or stencil). In a similar technique to what pastrie chefs use to create "spritz biscuits" (moulded cookies, piped biscuits), here the material is either warmed up or heated and then pressed through a static or rolling forming duct. Although many materials can be extruded (plastics, elastomeres and metals), the process is mostly used for aluminum and aluminum alloys to create shapes and profiles difficult to achieve in other processes. In rack and computer equipment, extruded aluminum is very common, as backbone structures in enclosure designs, corner profiles of boxes and as construction aluminum square bars, for example like solar panel mounts. It is also used to fabricate heat sinks and other smaller components where an extruded metal profile is cut down to result in a desired part. Rack mounting rails without pre-defined mounting positions are available as extruded aluminum profiles. Also, popular structural framing systems "T-Slot" and "V-Slot" use extruded aluminum profiles. See T-Slot for more.

Fan tray

In closed rack cabinets, thermal management is crucial. A fan tray is a horizontal, usually 1U or 2U high rack module with horizontally installed cooling fans, covering the whole surface of the module. Installed as part of a rack assembly, a fan tray can help with in-cabinet ventilation or, when installed at the top or bottom, to exhaust hot air, or take cool air in.

Fixed Rails

The alternative to sliding rack rails are fixed rails. Essentially a fixed rails is like a rack shelf, except it is missing the middle horizontal shelf part (the equipment will form the middle supporting part, once installed). A fixed rail consists of an L-shaped profile that is either as long as the whole rack is deep, extending all the way from the front post to the rear post, or a set of two half-length brackets, that are fixed on the front and rear of a two-post rack. Once installed, the L-bracket (or the set of two rail brackets) forms a small shelf on either side of the rack, allowing the edge of equipment to rest on these brackets in a non-bolted fashion, allowing equipment to be slid into or out of the rack slot, grinding over the smooth surface of the L-profile. One downside of fixed rails is that once a server or chassis is pulled, it has to be pulled out completely and placed somehwere else as the fixed rail support is only acting like a table and pulled equipment may become front-heavy or otherwise unstable unless it is completely resting on the L-bracket. One upside is that fixed rails can be variable in length, and thus be adapted to variable depth rack frames or rack cabinets - sliding rack rails are usually fixed length due to how their roller meachnism is laid out. Also compare Fixed rack shelf.


metal thickness in the United States / North America is often described with the term "Gauge" (sometimes "Gage"), a measure unit used for wires, screws and raw sheet metal. Abbreviations are either a capital "G", "gg" or "Ga.". Its name originates from the "measuring gauge", a tool of the same name to determine the thickness of wires. Gauges for wires are usually round, with round holes to measure wire diameters, and gauges for sheet metal are usual rectangular, resembling a ruler, with rectangular notches on one side for mesuring. Standards like the US "American Wire Gauge" (AWG) or the UK "Imperial Standard Wire Gauge" are still widely used metrics to classify metal wire thickness. From its origin in round wires, the term Gauge is similarly used to describe the diameter of hollow needles / cannula in medicine, the diameter of musical instrument strings, the diameter of jewellery or the caliber of ammunition. One important difference to other metering systems is that it's reversed in such a way that larger numbers describe thinner material thicknesses or smaller diameters. Further, a Gauge rating is related to the type of metal described, meaning, for example, "16 Gauge steel" is 1.519mm thick (standard steel), while "16 Gauge galvanized steel" is 1.613mm thick, 1.588mm with stainless steel and 1.29mm with aluminum, brass and copper. Gauge sizes don't follow some defined formular, thus conversion from Gauge to Millimeter or from Gauge to In ches has to be done with a (appropriate for the measured material) lookup table. It is common to only use every second Gauge value, for example 22 Gauge, 20 Gauge, 18 Gauge, 16 Gauge, etc. Also, the range of common Gauge values varies between measured materials, for example aluminum is measured from 40 Gauge to 6x zero Gauge thickness (14.732mm).

Gauge conversion table

Standard Steel Galvanized Steel Aluminum
Gauge in mm in mm in mm
10 0.1345 3,416 0.1382 3,51 0.1019 2,588
11 0.1196 3,038 0.1233 3,132 0.0907 2,304
12 0.1046 2,657 0.1084 2,753 0.0808 2,052
13 0.0897 2,278 0.0934 2,372 0.0720 1,829
14 0.0747 1,897 0.0785 1,994 0.0641 1,628
15 0.0673 1,709 0.0710 1,803 0.0571 1,45
16 0.0598 1,519 0.0635 1,613 0.0508 1,29
17 0.0538 1,367 0.0575 1,461 0.0453 1,151
18 0.0478 1,214 0.0516 1,311 0.0403 1,024
19 0.0418 1,062 0.0456 1,158 0.0359 0,912
20 0.0359 0,912 0.0396 1,006 0.0320 0,813
21 0.0329 0,836 0.0366 0,93 0.0285 0,724
22 0.0299 0,759 0.0336 0,853 0.0253 0,643
23 0.0269 0,683 0.0306 0,777 0.0226 0,574
24 0.0239 0,607 0.0276 0,701 0.0201 0,511
25 0.0209 0,531 0.0247 0,627 0.0179 0,455
26 0.0179 0,455 0.0217 0,551 0.0159 0,404
27 0.0164 0,417 0.0202 0,513 0.0142 0,361
28 0.0149 0,378 0.0187 0,475 0.0126 0,32
29 0.0135 0,343 0.0172 0,437 0.0113 0,287
30 0.0120 0,305 0.0157 0,399 0.0100 0,254
31 0.0105 0,267 0.0142 0,361 0.0089 0,226
32 0.0097 0,246 0.0134 0,34 0.0080 0,203
33 0.0090 0,229
34 0.0082 0,208


is a proprietary coating process and coated sheet steel brand, consiting of 55% aluminum, 43.4% zinc and 1.6% silicone. Galvalume is a registered trademark of BIEC International, Inc. Galvalume is a highly effective anti-corrosion process and a popular material choice in rack equipment.

Galvaneal Steel

or "galvannealed steel" is steel that has been galvanized and then annealed in a furnace as a subsequent second step. The furnace is used to heat the metal above its recrystallization temperature, to over 500 degrees Celsius, where it alters the physical properties of the material. Annealing ("glowing") produces a more workable metal, it's not as hard anymore and less brittle, easier spot weldable and can be easily painted (vs. galvanized steel, which may be hard to paint, depending on process) at the expense of a lowered corrosion resistance in comparison with galvanized steel. Galvaneal Steel of 18 Ga. thickness is a common material for open frame telco racks, as racks are usually painted and not as strongly exposed to environmental influences.


Galvanization or "galvanizing" is one form of zinc coating and part of a number of processes of protecting metals against corrosion. The process applies (lays, in a process of "plating") a thin zinc coat onto a metal surface, usually by submerging parts or a whole structure into a bath of zinc electrolyte ("zinc plating", "zinc coating"). This is similar to Dip-Coating as it allows the zinc to set evenly and/or reach every unevenness or crevice of the processed workpiece. Similarly as ECD in Dip-Coating, an electrolytic charge is used to optimize the particle deposition on the bathed object ("electroplating", "electrogalvanization") in a "electro-chemical deposition" (ECD), which is part of an overarching category of coating processes, "chemical solution deposition" (CSD). The result is a fine textured corrosion resistant surface well suited for painting or powder coating. In comparison, the "Hot-dip galvanization" ("hot zinc dip", "Feuerverzinken") process, where a workpiece is lowered into a bath of molten zinc and no electric charge is applied, forms a zinc surface that is prone to oxidization and thus hard to paint. Hot zinc dipped steel can be distinguished by its typical surface, a crystalline pattern, resembling Voronoi cells, where each cell reflects light slightly different ("flowering").

Ground Loop

with racked (analog) electrical equipment, in audio and video, low frequency hum noise or more general extraneous noise can sneak into systems by having multiple connections/paths (ground conflict) to a single electrical ground (earth), leading to potential differences between devices. The actual "hum" is the sinusoidal wave produced by 50Hz or 60Hz mains power. "Hum loops" are particularly of concern in audio, where a ground-loop or "hum loop" would deteriorate the audio quality of sound-producing equipment. It is very common with amplifiers, like guitar amplifiers or analog audio effect gear. While "hum" in analog connections is quite easy to spot (hear or see), the same type of noise does impair digital signal connections as well, leading to drop-outs or similar on the digital path. To mitigate hum, racked equipment needs to be properly grounded, that means grounded via the path the equipment manufacturer intended. With systems equipped with a three-pole IEC power inlet, one of the poles is for ground. As racked modules usually also will make conductive contact with the metal mounting rails of a rack via the mounting front-plate and screws, it has to be decided if a module is having an unintentional second ground contact. Using isolated screws, taping, plastic spacers and similar to isolate the front-plate from a rack might break this second ground path and eliminate a ground-loop. having a wooden rack stand, which is a non-copnductive material would have the same effect. Other rack equipment is expected to make contact with ground via the front-plate, so this needs to be investigated on a case by case basis. Plugging two devices from the same rack into two different power outlets / wall sockets will often lead to ground-loops - where connecting both devices to the same power strip with the same ground potential will often solve the issue. Another (common but dangerous) approach is to cut ground via the power cord of one module (with a device called a "ground lifter" or "cheater plug") and accept the fact that one device is making ground contact somewhere else. But it must be said that this "somewhere else" can't be guaranteed and disconnecting ground in order to mitigate hum can be a serious safety issue. With long cabling, as common in audio or video productions, potential differences may also exist between different rooms or buildings, leading to noise problems. For such cases, dedicated "hum eliminators" can be used or cable ways may be isolated by using optical fibre connections. Having proper and reliable ground for electrical equipment is essential to eliminate noise and for safety reasons. Not every power outlet can be trusted in terms of proper wiring and providing proper ground. A device called a "Ground Fault Tester" (GFT, or "Earth Ground Tester" or "Ground Fault Sensor") can be used to test a power outlet and make sure it provides ground. Additionally, an outlet needs to be wired properly in how the hot, neutral and ground wire are laid out. Having miswiring here will lead to stated problems and safety issues. In audio and music, getting a small electric shock while connecting equipment, like guitars, or from a microphone's metal surface, is a clear indication for massive grounding issues. With computer equipment, microelectronics and Electrostatic Discharge Sensitive Devices (ESDS) proper ground is not only required to eliminate mains hum but to mitigate any built-up of even the smallest charges. Compare the section on "ESD" for more.

Half-rack format

Comparable to 19-inch rack frame enclosures, a half-rack, 10-inch, or 9.5" rack only uses half the width of a standard 19" rack, thus offering a modular mounting option on a smaller footprint. Similarly as there is no norm or standard for the half-rack format, the term is somewhat fuzzy. It may refer to equipment being half wide, but may just as well, in different contexts, be used to describe equipment that is only half-deep. Especially with networking equipment, it is common that devices themselves aren't very deep but are installed in deep racks, back to back, so that two devices can share a 1U or 2U slot in a four-post open rack or rack cabinet. Half-rack then may just as well describe a rack or rack-cabinet that is only half the size of the very common 42U "full height" rack format. When the term half-rack is refering to a device that is only half the size of a normal 19" enclosure, it further is not clear if this means a device that would fit the 19" rack envelope on its own or if the device itself is even smaller, allowing some chassis or shelf system to carry the device and still fit into the 17.75" width of the 19" rack aperture. One exemplary audio half-rack device is 218mm wide, 215.5mm deep and a 44.35mm tall front plate (the main body of the device, excluding optional rubber feet, is only 42.0mm tall, a special rack shelf with a 20mm recessed front, fitting behind the slightly larger front of the device, makes sure the system does not exceed the defined max. height for a 1U rack module). Compare 10" rack for another "half rack" industry standard.

Hard reset

turning a device off and on again, in order to trigger a reboot or restart. On uptime optimized devices, this is usually the last resort when an otherwise unrecoverable situation arises. Normally, system administrators would prefer a "soft reset", where a system is going to halt/reboot/restart via defined procedures, usually triggered in software from the local or remote shell/console.

Harting Connector

Harting is a multinational manufacturer of electric connectors and the "Han" line and similar connectors made by Harting have become synonymous for multi-pin rectangular macro connectors that are popular in the (live) event, show and audio space to connect flight case racks and subracks via thick Cable assemblies, usually colloquially called "cable snakes" in the audio industry. (Note that a "Subrack" here describes a smaller, trolley type audio 19" rack of around 21U height, not IT Subracks for vertical cards.) The 16 pin Harting connector (dimensions 109.60 x 35.50 mm) is a connector with screwing terminals inside the rectangular outer shape. Other Hartings come in 48x35mm, 65x35mm, 82x35mm, 86x71mm or 108x35mm and have screwing outside the cutout rectangle of the connector. The large Harting connectors are usually rated for multi amp power applications, as needed for dimmers and connected lights. As part of cable managment, it is common to guide the thick cabling, in ducts etc. through a "Harting trumpet", a funnel-shaped (cone) hole protector that is intended to prevent wear and abrasion on the cable.

High Density

In datacenter design, "high density" means that equipment isn't "scaled out" but "scaled up", a common concept describing that the same or a growing amount of computing power is achieved not by letting the whole system grow in size but by increasing the density within the same or a smaller space. For the computing world, this means that the consumed wattage in a single rack increases. While a few years ago, a common metric for a single rack was to consume around 2000 - 4000 Watts of power, a high density rack of today can easily consume between 10 or 15 kilo Watts of power. In High Performance Computing (HPC), this can even go up to 80 - 200kW of power per rack. The CPU density, or more general, the compute density per rack unit and with it the density per rack has sharply increased. Virtualization increases system utilization. Smaller technologies, over blade centers to microservers, have made the amount of flops per square foot skyrocket. High density GPU processing as common in image rendenring or artificial intelligence applications is taking up more and more space in today's datacenters. The Open Compute Project is one of many initiatives to drive a high density mindset into new datacenter designs. And with a higher density in physical hardware and higher thermal dissipation, traditional infrastructure of datacenters came to its limits. New facilities usually opt for slab flooring, as racks can easily weigh over 2 tons (4000 lbs.). New approaches in airflow and cooling management moved air ducts and rack cabling to overhead ducts and cable raceways. High density bears many challenges, but it forces the ecosystem to invest in efficiency, and, for example, reduces the overall footprint and reduces cable lengths. High density is part of the general miniaturization of computing equipment all around.

Hifi Rack

a Hifi Rack (also "Hi-Fi Rack") is a type of audio furniture used to carry (stacked) components of a Hifi system (or "home audio system"). Home audio systems in their earlier incarnation around 1940, 50s and 1960s used to be monolithic systems in the format of wooden commodes or cabinets, with turntable, tuner and integrated speakers. Beginning around 1970s and 1980s, the elements of these integrated systems became separated when listeners tried to optimize the sonic quality of individual components by mixing gear from different manufacturers. Component based home audio systems became the norm and a staple of many households for a number of decades. A common setup of audio components is Amplifier, either dedicated or integrated with a Tuner as a "Receiver", a dedicated Tuner/Radio and a mix of media playback devices, a Turntable, Tape Deck or CD-Player and sometimes an additional audio Equalizer for tone adjustment. Home theater systems also employ Receivers capable of decoding multi-channel audio and do video switching. More recent components usually offer streaming audio network connectivity and USB or smartphone docking. Audiophile users, ever striving for improved audio fidelity, often do not simply stack individual components on each other, but use dedicated audio furniture, Hifi Racks or A/V Racks, to stack components. These Hifi Racks usually offer only one shelf level for one component and promise to isolate audio components from from each other's and environmental energy, vibration and noise - resulting in superior acoustic performance of the system's components.

What is the origin of Hi-Fi audio component sizes?

Home audio Hi-Fi system components usually follow a de-facto industry form-factor, so that when listeners mix and match components from different vendors, the system as a whole would still be stackable neatly as one system. As early audio systems were simply electric instruments, engineers opted for popular 19 inch rack dimensions. Also, some audio components would be manufactured so they could be either stacked stand-alone or installed in a screwed rack with the help of rack-ears or on a rack shelf. So in result, most home audio components would fit into the rack opening of a 19" rack, which is at max. 450.85mm or 17.75" wide. Every few years, a new attempt to establish smaller or "mini" component footprints was made. Pioneer, Sony, Onkyo and others released multiple generations of devices of 2/3 rack width, being around 320mm wide or even less wide.

Exemplary dimensions of Hi-Fi audio components

  • Kenwood KA-701, a 1980 stereo amplifier, is 440mm x 153mm x 407mm (W/H/D) at 13.5kg
  • Krell K-300i, a 2024 high-end integrated stereo amplifier, is 438 mm x 105 mm x 457 mm (W/H/D) at 23.6kg
  • Pioneer SX-N30AED-S, a 2024 Stereo Network Receiver, is a 435mm x 149mm x 327 mm (W/H/D) at 8.3kg/18.3lbs.
  • Onkyo TX-SV 9041, a 1994 A/V-Receiver, is 455mm x 170mm x 388mm (W/H/D) at 13.5kg and 1990s Onkyo components generally being a little wider than usual
  • Yamaha AX-590, a 1995 stereo amplifier, is 435mm x 146mm x 389mm (W/H/D) at 10.4kg
  • Technics SL1200MkII, a high-end turntable, is 453mm x 162mm x 360mm (W/H/D) at 12.5kg
  • Nakamichi RX-202, a high-end auto-reverse tape deck, is 450mm x 136mm x 255mm (W/H/D)
  • Pioneer CT-90 R, a 1983 single tape deck, is 420mm x 120mm x 355mm (W/H/D) at 7.2 kg
  • Tascam 238 S, a 1994 8-track tape deck, is 482mm x 149mm x 355mm (W/H/D) including detachable rack ears, so it's 19" rack compliant
  • TEAC R10, a 1990 high-end DAT recorder/player, is 442mm x 150mm x 355mm (W/H/D)
  • Sony MDS-JA 20 ES, a 1997 MiniDisc recorder/player, 430mm x 125mm x 345mm (W/H/D) at 7.8kg
  • Pioneer LD-S 1, a 1986 LaserDisc Player, is 457mm x 136mm x 468mm (W/H/D) at 16.8kg
  • JVC XL-V 221, a 1990 mid-range CD-Player, is 435mm x 92mm x 290mm (W/H/D)
  • Pioneer CDJ-500S, a 1997 early top-loading DJing CD-Player, uses a half-rack format, being 217mm wide, 217mm x 98mm x 228mm (W/H/D) while the later iconic Pioneer CDJs are usually around 320mm x 98mm x 360mm
  • Denon DN 2000F, a 1990s pitch-adjustable professional dual DJ CD-Player, equipped with rack-ears for rack-mounting, is built to fit into a standard 19" rack

Home mounts

Not really a term from the enterprise or datacenter rack world, yet it describes mounting solutions for the Small or Home Office (SoHo) environment, offering to streamline cable management, put away with tangled cable nests and to upgrade computer setups through clever wall, deskside or under-desk fasteners.


common abbreviation for "horizontal pitch", the unit of measure for horizontally laid out front-plates/components in 10" or 19" Subrack rack positions. More on vertical cards usually mounted side by side under (cmp.) "Subrack".

In-Cabinet Rack

With edge installations, branch networks or smaller IT deployments, it is common that computer hardware is installed in a (sometimes customer-facing) standard office environment. In order to hide the technical hardware and achieve a less industrial appearance, it is common to have 19" racks installed inside wooden furniture cabinets. Some racks are especially designed for this practice and feature special bolting options, slide or swivel out mechanisms and other ergonomic features to allow technicians easy access to cabling and equipment.

Inches to Decimals to Millimeters conversion table

inches decimals mm
1/8 0,125 3,175
1/4 0,25 6,35
3/8 0,375 9,525
1/2 0,5 12,7
5/8 0,625 15,875
3/4 0,75 19,05
7/8 0,875 22,225
1/16 0,0625 1,5875
3/16 0,1875 4,7625
5/16 0,3125 7,9375
7/16 0,4375 11,1125
9/16 0,5625 14,2875
11/16 0,6875 17,4625
13/16 0,8125 20,6375
15/16 0,9375 23,8125
1/32 0,03125 0,79375
3/32 0,09375 2,38125
5/32 0,15625 3,96875
7/32 0,21875 5,55625
9/32 0,28125 7,14375
11/32 0,34375 8,73125
13/32 0,40625 10,31875
15/32 0,46875 11,90625
17/32 0,53125 13,49375
19/32 0,59375 15,08125
21/32 0,65625 16,66875
23/32 0,71875 18,25625
25/32 0,78125 19,84375
27/32 0,84375 21,43125
29/32 0,90625 23,01875
31/32 0,96875 24,60625


short for "Intelligent Platform Management Interface (IPMI)" is a suite of interfaces used to remotely monitor a host system. IPMI is an out-of-band (meaning a separate system) administration tool. In IPMI a "baseboard management controller" (BMC) is a complete but separate system attached to a host, a microcontroller or low-spec computer system, that acts as a controlling instance of the monitored host (especially servers), offers communication, own serial and network ports, etc.


Some enclosures feature pre-cut holes where the cut is done as a perforation of the enclosure's material, either sheet metal or (abs) plastic. A knockout can then be pushed through with manual force only when a cable duct or hole is needed. A related idea is to have "gland plates", sub-sections of a door or side-panel that is often made from a different material, having pre-cut openings ("grommets") for cables to be run through, being padded to prevent damage to cables and sometimes feature a brush curtain to seal unoccupied space, helping to mitigate unwanted airflow.

KVM Mount

Maintenance Bypass (MBP, MAINS mode)

also known as "Service Bypass" is a fault-tolerance and servicing feature of some Power Distribution Units (PDUs). Without shutting down the equipment, the power can be switched away from the UPS when it needs to be replaced. It is common to serially link the main power to a UPS first and then feed UPS -backed power to a PDU. In case the UPS in this setup fails, it wouldn't be possible to replace the UPS without interrrupting power to the PDU and with it to connected rack devices. For this scenario, some PDUs have an additional power inlet where mains power can be connected so that connected equipment can draw input power from the MBP through a dedicated power inlet.

Lace Bar

part of wire management (cable management), a "lace bar" is a vertical or horizontal support element, that helps to route and support in-rack cabling.

Laser cutting

Laser sheet metal cutting is a common technique in manufacturing shaped workpieces from raw sheet metal blanks. Laser cutting belongs to a category of techniques to cut material with heat, like flame cutting ("oxygen torch cutting") or plasma machining. Also, laser cutting is an umbrella term for a number of methods where different types of lasers are used to produce different types of separation within a worpiece, like vaporization, melt and blow or reactive cutting using a reactive gas and a laser as ignition source. Laser cutting is slower than cutting metal pieces from sheets in stamping presses but more flexible with less tooling required and more flexibility. Apart from flat sheet metal laser cutting, there are dedicated machines for laser pipe cutting. Today, laser cutting machines are often embodied as "combination machines" where a laser turret is able to cut arbitrary shapes from a metal sheet and an attached bending mechanism, either a press brake or cornice folding mechanism, is able to bend workpieces according to computer controlled shapes. Laser cutting is an alternative to (cmp.) "water jet cutting". The groove created by a laser cut ("kerf") is very thin, usually thinner than with waterjet, yet with laser cutting there's a thermal impact on the cut material and this may change the material's structure near the kerf. Depending on the used method, the cut surface can be of high smoothness or of less perfect optical quality (usually when oxygen is used in reactive cutting), requiring a intermediary step of deburring edges prior to e.g. powder coating.

Laser cutting machines manufacturers

Laser welding

is a metal welding technique where a laser beam is used to inject heat energy into the welded area. As the beam is able to melt the material at the weld, usually no additional consumable is required to fuse the workpiece material and produce a welding seam. Laser welding is a relatively new technique and mostly applicable for thinner metal around 1mm thickness and below. The noble gas argon is usually used as shielding gas during the process. In comparison with traditional MIG/MAG welding, laser welding has a number of advantages (depending on application), like less or no thermic warping, high surface quality and easier handling. That said, it is also usually more expensive.

Metal forming

To produce workpieces of high strength and precision, metal is one of the best materials available. In metal forming processes, metal can be brought into many shapes, by pressing it through a die in extrusion, or applying a forming force, as is done in metal presses or metal bending machines. Blanks of thin sheet metal blanks can be formed by being bent on bending machines, either folded (flanged) or clamped (in a press brake). Stamping presses are (usually) large format hydraulic or servo driven steel presses (or other metals) where a die is used to deform, indent or cut flat metal blanks ("hard tool stamping"). Punching machines work on sheet metal and punch holes into metal blanks, like a cookie cutter.

MIG/MAG welding

is a common abbreviation for the traditional way of welding metal, part of the category of "Gas metal arc welding" (GMAW, also "GTAW" for "Gas tungsten arc welding") techniques. The abbreviation "MIG" here is short for "Metal inert gas" and "MAG for "Metal active gas". In the process, a strong current is applied to a wire electrode, forming an electric arc between the electrode wire of the welding gun, melting the consumable wire and the workpiece at the welded spot, forming a fuse of the materials at the welding seam. During the process, a noble shielding gas, usually argon or helium, is injected into the weld area to protect it from oxygen or water vapour contamination. being the "old way of welding, MIG/MAG in comparison with the more modern laser welding has a number of drawbacks (depending on application): a lot of thermal energy is injected into the weld area, making workpieces prone to thermic warping, the process requires preparation, is complex, consumes and mixes materials and produces less perfect weld seams. That said, it is cheaper and the only option for workpieces > 1mm thick. The innovative process of Cold Metal Transfer (CMT) welding, developed by Fronius, reduces the amount of thermal energy during a weld and allows to reduce warping or burn throughs especially on thin sheet metal.

Monitor Mount

With many systems in one rack, it's common to have multiple machines share one keyboard and one monitor via switching hardware. Many monitor and/or keyboard mounts are embodied as drawers, so when an operator works in front of the rack, it can be easily slid out, or stowed away when not in use.


abbreviation for "Numerical Control" or "Numerical Control Units" (NCU). Historically, the "numerical control" was an innovative add-on to existing manufacturing machines, an electronic control unit that was able to control a machine according to a specific program ("NC Program"), a succession of guidance commands ("G-Code"), allowing to machine parts to specific specifications and at the same time, being able to easily reprogram a machine for a different part with different specifications. NCs were introduced during the 1940s in the United States and were dedicated electronic circuitry to process program input (usually via punched cards) and control connected motors or servo mechanisms. It was during the 1970 and 1980s that these electronics were improved with or replaced by then "mini" or "micro computers", making a "NC" a "CNC" - a "Computerized Numerical Control". Today, the term CNC and NC is mostly used interchangeably - with local lingo and preference varying throughout the world - while CNCs have completely replaced traditional NC systems. Turning lathes today are often CNC machines - especially when incorporated as multi axis milling machines - or automated punching machines ("numerically controlled turret", "NCT") or laser-punching machine combinations. The tradition of regarding the control unit of a machine as the "NC" lives on in modern products like the "SINUMERIK 840", a machine control unit manufactured by Siemens that is, for example, used in Trumpf TruPunch machines and today part of the "Simatik SPS" line of "Programmable Logic Controllers".


NCT is the abbreviation for "numeric controlled turret" or "number controlled turret" or "CNC turret punching machines". While the more general category of turret punch Presses/ punching machines may also be manually operated, a "numerically controlled turret" is a type of automatic and computer controlled machine, similar to CNC turning lathes or multi axis CNC machines, where the stamping tool is guided by computer input. "NCT piercing" describes stamping an arbitrary form or shape into or through sheet metal blanks. "NC stamping" is usually regarded as an alternative (more flexible, more economical) to large format stamping presses, as the latter require more preparation, a tooling investment, where a die needs to be machined etc. On the contrary, an NCT machine isn't able to perform all forming techniques and is slower and less strong. The difference between a NCT machine and a machine stamping press is similar to that of a typewriter or an inkjet printer vs. an (offset) printing press. NCT is an older term for today's automated punching machines and seems to be more widespread in use in Asia - for this reason, more can be read under Punching machines.

Notching machine

a notching machine is a simpler (usually manually operated) type of Punching machine and used to form metal. A hydraulic, pneumatic or manually operated turret with a stroking tool is used to "nibble" notches ("corner notching machine") or arbitrary shapes away from the edge of inserted sheet metal. Some machines are able to bend or deform blanks in such a way that a seamless (no weling required) corner cavity of a sheet metal pan is created ("corner former") - a simple form of forging.

Nut Tool

Open Compute Project

A collaboratively industry workgroup initiated by Facebook to redesign major components of datacenter infrastructure, with the aim to increase efficiency, reduce costs and improve overall management of basic datacenter building blocks. Some central topics are cooling and power distribution, physical arrangement of systems in redesigned Server Racks and reusable server blade enclosure design.

Open Frame Rack

Open Rack

(as designed by the OCP)


Linux Foundation originated open-source implementation of a "baseboard management controller" (BMC) firmware stack.

Panel bender

is a sheet metal bending machine. Panel bender is antoher name for a "Cornice brake", machines that produce a bend by bending the metal against an edge to produce a radius ("swivel bending"). Results are similar to what a "Press brake" does but the mechanism is different.


as a chemical process usually means the intended coating (some metals form spontaneous surface oxidization) of a (metal) surface in order to mitigate corrosion, from oxidization or other reations with the environment. Sheet metal, as frequently used in rack applications, is usually passivized in one form or another. One simple way of passiviation is painting. Aluminum is either processed through chromate conversion coating or anodized. Copper or silver is often gold plated or zinc-nickel plated to prevent corrosion.

Patch Cable

a form of shorter cable (usually around 5-20 inches, 10-60cm) used to connect electronic, opto-electronic or optical devices with each other. A patch cable is sometimes known as "patch lead", "patch cord". It is common to have colored patch cables to color-code their use and prevent cabling faults during installation or maintenance.

Patch Cables in Audio, Music and Sound Recording

Patch cables in audio technology are usually wires with either two or three wires and a cylindrically-shaped "phone connector on either end. Patch cables with two wires use "TS" (Tip Sleeve) connectors with "two conductive rings", and cables with three wires "TRS" (Tip Ring Sleeve) plugs ("TRS phone connector") having "three metallic rings". Such patch cables were common in telephone switchboards of the POTS (plain old telephone system), jack fields (patch fields) where operators were able to patch phone connections/conversations physically together. The name stuck. Today, patch cables are common in the audio recording field where rack-based "patch bays" are used to route audio signals from one device to another, in a flexible way. For example to route a select set of tone generators to a limited number of recording inputs. With modular synthesizers, patch cables are used to "patch" individual components of the tone-generation path together. When routing audio signals, usually three-wire patch cables are used to transmit a "balanced" mono audio signal. "Balanced audio" is less susceptible to external noise caused by electromagnetic interference on longer cables. Although patch cables are short, they usually are part of a longer cable path and with the patch cable being one segment of this path, requires patch cabling to be three-wired as well.

Patch Cables in Computer Networking

Depending on what technology is employed, in computer networking a patch cable may refer to cable with either an optical or electrical inner technology. Fiber-optic cables use one or more optical fibers to carry light. Electrical cables use an inner core of one or more conductive wires made of copper or another cunductive alloy with a link impedance of 100 Ω, usually multiple twisted-pair copper interconnects. Network patch cables today use shielded or unshielded (ISO/IEC 11801) Cat5 ("Category 5"), Cat5e, Cat6, Cat6a, and as of 2024, increasingly Cat 7/7a and Cat 8/8.1/8.2 cables for high speed networking. Cables are terminated using 8P8C (RJ-45) modular connectors with straight-through T568A or T568B wiring.

Patch cables in TV and Video

the SMTPE (The Society of Motion Picture and Television Engineers) standardized "Serial digital interface" (SDI) is a technology for sending video signals over coaxial cables. SDI and HD-SDI use coaxial cables terminated with BNC connectors and are commonly connected via Coaxial or BNC patch bays.

Patching in fiber optics

In fiber-optic networking and communication, jack fields or patch fields are usually called "Splice Box" or "Splice Drawer" and are part of "fiber management" (the equivalent of cable management, just for fiber-optic cables). A splice drawer is used to accommodate and protect fiber splices and their characteristic short loops in 1U or 2u height. Very often, these splice compartments are embodied as drawers, as, in contrary to electrical patch jacks, fiber optic splices must be laid flat and twist-free inside such a drawer.

What material are PC cases made out of?

Computer enclosures come in a variety of sizes and forms, even the term used to refer to these boxes illustrates the variety: When professionals speak of industrial or data center grade systems, usually the term "chassis", "server chassis", "computer chassis" is used. In private everyday language, when people refer to the outer shell of a computer, they speak of a "personal computer case", a "PC case". Although a PC case can be made of any material, plastic, wood, acrylic, etc., the most common materials are steel and aluminum. A run-of-the-mill "beige box" desktop or tower (which is usually black since the turn of the century) is commonly made from 0.75 - 1.2mm electrogalvanized steel - bent and cut into shape in large stamping presses, then riveted and welded, optionally painted. This is for large volume produced PC cases. For specific use cases and with servers, the volume of produced cases may be smaller and the processes more boutique-like. Metal may then be laser cut and bent on manual or semi-automatic press-brakes instead of stamping presses. Some or all parts may be from aluminum instead of steel, may be of higher thickness and corrosion mitigation may be achieved differently. Where weight is a factor, like in mobile applications or airborne installations, all-aluminum enclosures are common. With applications where specific Electromagnetic compatibility (EMC) is required, materials are specifically coated and the design takes shielding and sealing of the enclosure into account. With many modern domestic PC cases, where large glass or acrylic areas expose the internals of a system, show off RGB fans or illuminated memory bricks, the actual shielding of generated high frequency EM radiation is often woefully neglect.

PC Rack Mount

In corporate IT deployments, oftentimes off-the-shelf PC systems are used as (cheaper) building blocks of computer infrastructure. It is common in web hosting, computer cluster research or small (or SoHO) creative studios to use "generic beige boxes" on Rack Shelf in a High Density setup. Some tower workstation PCs are designed to fit horizontally into the rack opening of a standard EIA rack, and can be screwed in with the addition of Rack Ears. Some PCs are designed to fit side by side into a rack's opening, in groups of two, three or four machines, in 2U or 3U of vertical space. The Cobalt Cube server appliance sold during the 1990s was popularly stored this way.

Micropolis at one point offered the Raidion family of desk-side storage towers. The 3.5-inch drive version Raidion LS included the option to store the usually vertically laid out tower on-the-side horizontally in a server rack with the help of a shelf and mounting brackets, offering customers to move the appliance into a datacenter.

PDU (Power Distribution Unit)

is a multi-outlet power strip device, designed to distribute power to multiple devices (within a rack, "Rack PDU") from a single power source, where the power source might be the utility grid, a UPS or a generator. In the field of audio, a PDU is sometimes called a "rack rider" as it is common to install the PDU on top of audio racks, with utility lighting for devices below and power regulation (power conditioning) built-in, to protect rack equipment from power surges, etc. In computer datacenter applications, PDUs are sophisticated pieces of hardware, enabled to not only distribute power but also power-cycle connected equipment, meter power draw, measure room temperature and humidity and transmit such metrics to remote logging and monitoring equipment (cmp. "DCIM"). Power capacities of PDUs are usually rates in amperage. PDUs are available in standard horizontal rackmount designs of 1U or 2U and often optional vertical installation (of 0U, "zero U", "ZeroU"). While all PDUs provide basic power distribution, some models allow local or remote monitoring, on the PDU unit level or down to PDU outlet level, with outlets being metered, switched and metered or the whole PDU ("Smart PDU", "intelligent PDU", "iPDU"). Power redundancy on the PDU level might be implemented via ATS (Automatic Transfer Switch), where the PDU uses multiple power sources and switches from one to the other in case of failure of one (power failover). PDU come in localized for, with output receptacles suitable for the region of the connected devices. Some PDUs offer "Maintenance Bypass" (MBP). Some PDUs offer "Branch Circuit Protection" with a given group of outlets having their own fuse and/or circuit breaker. Some PDUs support the use of "3-Phase power", with phase indicators on PDU outlets.
Power Distribution Unit (PDU) cabling options: using height units, 1U or 2U; or vertical zero U layout along the side, saving rack space
Power Distribution Unit (PDU) cabling options. On the left: a PSU occupying height units, 1U or 2U; on the right: PDU in vertical "Zero U" layout, along the side, saving rack space.

Portable Rack

Powder coating

aluminum, steel or more general metals are all prone to changing their surface in oxidation processes (corrosion), a generally undesirable effect that can be mitigated by applying a protective surface coat. Despite good adhesive properties of standard acryllic or enamel based varnishes, these lacquers often simply do not meet the requirements for durability and strength. Powder coating is a special type of (color) coating where the varnish, before application, isn't a solvent based liquid but a dry powder. This powder, either epoxy or polyester resin based, is applied with a powder gun and made to stick to the powdered object by an electrostatic charge. The powder coat is then cured by heating / baking the powdered object (the powder is "thermoset", "stove enamelling"). During the setting process of the coat, the coat forms a high-strength bond with the coated surface and builds up a scratch and impact resistant plastic surface on top, where the polymer chains have cross-linked with each other. Some newer powder coats can be cured by exposition to UV light, making the overall process less heat-intensive, allowing plastic objects to be powder coated. Powder coating may be one stage of coating, for example after a first step of dipcoat priming or before a finishing overpaint. Powder coats are available in various colors and types, with powders either creating a matte, textured or "rough" surface or cure to a shiny high-gloss finish. Computers, data center infrastructure and rack components are very commonly powder coated in matte or satin-matte coats of a black (RAL 9005), (light) grey (RAL 7035) or white color. For use in ESD aware environments, special Electrostatic Dissipative Coatings are available. ESD powder is mixed with a small amount of conductive additives, like metal. Common ESD additives are graphite, carbon black, carbon fibres, carbon nano tubes (CNT) or mica pigments (a silicate). Some of these powders limit the color palette to black. Some of these coats are more sensitive to abrasion.

Power Cable Plugs

Different power cord plugs, IEC 60320: C7, C13, C14, C19, C20
Different power cable plugs, according to norm IEC 60320: C7 female plug, as found on smaller A/V or IT equipment (SFF PCs, game consoles, etc.); C13 female plug and C14 male inlet socket often found on rack servers, networking equipment and other computer equipment; C19 female plug and C20 male socket are usually used as mains power couplers on PDUs and other power distribution equipment.

Power Conditioner

With racked audio gear, people often speak of "power conditioners" for one (less often two) rack unit high power distribution and power conditioning equipment, modules that commonly also feature rack illumination and a "central power switch" for the whole (sub) rack. See Power Distribution Unit (PDU) for more.

Power Whip (Whip/AC)

power cable assembly used to connect a data center rack to a the data center's or building's facility power busway / main power feed. These "whips" are used as an adapter, as different combinations of rack hardware use different types of connectors. Often a whole rack (via it's PDUs) or a rack's power shelf uses one type of connector, leading into a cable assembly with again a different type of connector, like a more universal AC plug. With an adapter to plug into 208V (30A/50A), 400V (32A) or 415V (30A) power outlets, a rack or parts of it can then, with such a "whip", be connected to a facility's mains power.

Power-Up Sequencing

feature of some PDU models, offering the ability to switch individual receptacles on, one after another, when the whole PDU is connected to power or switched on. This allows power to be be fed in sequence to connected devices, controlling the maximum amount of current flowing while systems power-on and boot or reboot, a situation where many systems draw higher currents than under normal load. Such a scheme is similar to what is known as Staged Spin-Up in data storage enclosures, where storage devices similarly are not spun-up all at once, to control "inrush current" - only here applied to a whole rack of devices and/or systems.

Press brake

A press brake resembles a stamping press as it is usually also embodied as a large H-frame assembly, with an upper element acting as a hammer and a lower element as an anvil. Inserted metal blanks, usually sheet metal, are bent "by delivering an accurate vertical force in a confined longitudinal area" (cited from the ASM tome on Press Brakes, see links at the end of this text). Press brakes are feedback-loop controlled and very accurate, driven either by mechanical, hydraulic or electro-hydraulic drives. The upper tool exerts a downward force into a V- or similarly shaped lower die to deform the workpiece. As the upper tool usually resembles a linear bar, the resulting bend is a straight line. When a worker or some computer controlled mechanism inserts the metal blank, dynamically positioned limit stops (backgauges) help position the sheet and control the length of the resulting flank. The top tools is usually easily exchangeable for specific bends. Computer controlled machines with automated feeder mechanisms are also often embodied as "combination machines", adding the ability to laser cut inserted workpieces.

Press brake manufacturers

PSU (Power Supply Unit)

a PSU converts mains AC line voltage into low-voltage regulated DC power, often provided in a number of low voltages at high currents, as required for the internal components of a computer. Early low voltage devices usually used linear power supplies, which use a rather large step-down transformer as a first step, where high voltage AC is converted into low voltage alternating current. A second step, a linear regulator implemented mainly with transistors then converts AC to DC. The drawback of this design is that a large portion of the converted power is dissipated as waste heat. During the 1970s and 1980s a new design emerged, made possible through advances in producing cheap transistors and control ICs, the "switching power supply". While a linear PSU also switches power, in sync with mains frequency, usually 50 or 60 times per second, a switching power supply switches several thousand times per second. In a switching PSU the incoming AC power is first converted to high-power DC power and then, similarly as in PWM, switched on and off to control the flow of power, resulting in a defined output power based on how long individual on-phases are. This scheme generates nearly no waste heat and switching PSUs may operate at 80-90% efficiency in practive, while a perfect design in theory should be able to achieve 100%. One side-effect of the switching mode is that it introduces high-frequency ripple ("noise") into circuits, requiring additional circuitry to filter out, while simple linear PSUs supply noise-free stable power. Today, computers and electronic devices usually use switch-mode PSUs. Many models also provide stand-by power, allowing a connected system to power down into suspended, "sleeping" hibernation modes, where software defined triggers are able to wake the system up, like "wake-on-LAN", "Wake-on-ring", BIOS or software-set timers or via Human-Interface-Devices (HID), "Keyboard Power ON" (KBPO). One further note is that, by design, switch-mode PSUs generate "common-mode" noise (aka "common-mode voltage" or "common-mode signal"), a high frequency alternating current on common ground. Personal-computer and server PSUs deviate this current towards ground via their "third pin" power plug ground connector (one more reason to have properly wired wall sockets), but two pin "wall-wart" style PSUs and many laptop PSUs do not. Common-mode noise is usually harmless for people, but it may interfere with or destroy sensible devices or circuitry.

Punching machine

Punching machines (mostly "Turret punch presses") work similar as much larger Stamping presses, as they use a flat bed to affix a sheet of metal and a (usually) overhead "stamp" that is driven vertically into the underlying metal surface to apply dents or cuts in the shape of the stamp ("ram" or "slide"), sometimes with a die below the stamped metal, working in tandem. Punching machines in general may be manually operated or automatic. The process of moving a sheet metal blank into the machine and maneuvering it around on the bed, below the stamping area, is usually servo and computer controlled on today's machines, which are also known as NCT presses, a somewhat older term. These machines are able to punch, deform, nibble, trim, shear, flare and flange the raw blank at a high pace. The turret is usually able to automatically exchange tools, sometimes in-sync with a die located on the underside of the punching bed. Wheel tools can be used to create rolling offsets and are a quick alternative to usually slower nibble forming operations. Some machine turrets are able to exchange complicated tools, like drilling mechanisms to cut screw threads. Modern machines are also often embodied as "combination machines", adding the ability to laser cut inserted workpieces.
NCT punching machine tool forms: a cup, ventilation louvers and bracket
Example punching tool forms of an NCT punching machine: a circular cup, angled ventilation louvers and a square bracket

Punching machine manufacturers (NCT)

Rack Adapters

Rack Box

Rack Converters

Rack Drawer

Rack ears

Equipment manufactured to be compatible with 19 inch (or 10 inch rack) form factors sometimes is shipped without the face-plate overhang to screw it into a rack. Such device enclosures are then designed as simple metal boxes with threaded holes on the side, so that additional "rack ears", or "rack mounting flanges" can be installed if required. Like any standard rack equipment front plate, rack ears commonly feature two unthreaded oblong holes for mounting, but might just as well have two horizontal notches on the rack ear edge, as notches were the original 1960s EIA standard. A rack ears option is usually chosen by device manufacturers when equipment is expected to be used primarily in "desktop mode", but may just as well be installed in a 19" rack. This configuration is common for audio devices, like tone generators, effects processors, etc. In A/V environments, a side-by-side mounting of two devices plus rack ears is an equally common setup. While most 19" rack ears are fitted to equipment that perfectly fills the rack opening of 19" racks, some narrower equipment might use special wider rack ears, acting as spacers to bridge the gap between the equipment side and the screwing strips of a 19" rack - for example when certain 500 Series audio housings are rack mounted.
Schematic view of 19 inch rack ears
Schematic view of 19 inch rack ears

Rack Handles

in data-center lingo, the term "rack handle" usually refers to the dor handle of rack cabinets. For security, many handles feature some kind of lock, ranging from simple key locks or padlocks to more sophisticated lock types, with smartcard readers, keycode field or wireless access mechanisms. In combination with mechanical door sensors, IR door sensors or other means of intrusion detection, rack handles are an obvious but effective way of implementing high-security or elevated physical security awareness for IT equipment.

Rack Hole Type

There are two types of rack holes in the vertical side rails of 10" and 19" racks, 12-24 Threaded holes and Square holes.

Rack mounting depth

There's no formal standardization for the depth of 19 inch racks ("mounting depth"), probably due to their development history from open frame structures with only two screwing posts.

Rack Power Strips

Rack Rails

also "slides" or "rack slides", sold as "rack rail kits". A common mounting helper rack accessoire that allows rack mounted equipmented to be easily slid out linearly for maintenance. Usually made from C-shaped formed zinc-plated or stainless steel rails (the "raceway" or "guide") with a roller carriage moving inside. Sometimes telescopic for an extraction ("opened") that is longer than the length of the guide in closed position. Rollers are usually ball-bearing rollers for less friction and a longer lifetime. Some designs combine roller and guide into integrated structures, sometimes made from milled or extruded aluminum. Some designs elaborate end-stops with damping systems to protect mounted equipment from shock or accidental rough handling. An important metric is the rated load capacity and the stiffness of the used material to minimize deflection in extracted state. The alternative to sliding rails - for example with constrained space, limited depth or when obsctructions prevent a real sliding rail to be installed - are Fixed Rails.

Rack Shelf

A rack shelf or "fixed rack shelf" is a simple horizontal support, usually made from sheet metal. A common embodiment is a three sided, folded structure with screwing terminals and an opening on the front, able to support arbitrary equipment inside the rack opening of a standard rack. As the horizontal element - the actual shelf - spanning the gap between the two front or four vertical posts need to be stable, some rack shelf feature an upward or downward folded lip that may obstruct equipment below or limit the usable opening height of the rack shelf
Typical generic 19-inch Rack Shelf
Typical generic 19-inch Rack Shelf (cantilever rack shelf)

Rack to Tower mounts

(Convert rack servers to tower servers, rack workstations to desk-side tower workstations)

Raised floor

In traditional datacenter design, floors were usually raised, meaning the walk-on floor is an artificial level created on stilts above the real, much firmer (concrete), floor surface. This raised floor creates ducts that could be used to hide cabling and direct chilled air into rack cabinets. As warm air naturally rises towards the ceiling, this provided a simple form of airflow management. The floor was made up of removable tiles, that could be lifted and rearranged, and rack cabinets would exactly fit into this floor tile pattern. In recent years, datacenters have seen a rise in density per rack and a sophistication of air conditioning, leading to higher required load capacities of raised flooring or floors in general and new approaches in cooling rendered the underfloor ducts less usable. Modern datacenter racks can weigh up to 2 tons (4000 lbs.) and exceed common flooring weight capabilities. That's why modern datacenter designs usually choose (concrete) slab flooring, able to support heavy equipment and allowing it to be bolted down (to counter seismic activity). This kind of datacenter relies on overhead air cooling ducts and overhead cable raceways to provide the infrastructure that was formerly (in parts) provided via raised floor and removable tiles.

Relay Rack

sometimes "2-post relay rack" or colloquially "post relay" is a type of rack with two free-standing vertical screwing strips. The name originated in TelCo installations, where it was common to install telephone sub-station relay banks into this type of rack.

Remote Management

Generic term to describe the (optional) feature of equipment or systems to be controlled remotely, from an off-site terminal or console.

Remote Power Management

ability of equipment, mostly PDUs, to switch power on a given receptacle in order to remotely power cycle servers or devices, to trigger remote reboots or recover from situations requiring a "hard reset".

What's the meaning of a stylized "RU"?

when you see a reversed capital "R" followed by a capital "U" ("backwards RU") then you've found the "Recognized Component Mark", a safety mark issued by UL Solutions, a US safety organization sometimes known as "Underwriters Laboratories". The "RU" mark is issued for components which are intended to be or become part of a UL certified product. As a mark exclusively intended for (electrical) components not end-products, it is usually not end-user facing. The RU mark is part of UL's "legacy UL Marks", a line of marks which UL separates from their newer "enhanced" and "smart" marks - yet the note of "legacy" does not mean that it has been retired or is not issued anymore.

Screw holes

Screws can be inserted into or inserted through metal in a number of ways and with different hole categories. Holes may be simple through holes or tapped for threaded bolts, may be countersunk or exectued as oblong slots. Countersunks screws have a conical head and are used with countersink holes. Screws with a cylindrical shoulder-type head are used with counterbore holes that feature an upper cylindrical part for the screw's head. A very shallow counterbore cylinder is called a "spotface" and is used with uneven surfaces or when a washer is used under a nut and a very aligned and even distribution of force is required.
Countersink hole, counterbore hole and through hole
Countersink hole, counterbore hole and through hole

Single Inlet Device

or a single-corded device is a kind of device that is not designed to have redundant power supplies and comes with only one cord - common with entry-level, consumer, pro-sumer or legacy computer and network equipment. A Power Distribution Unit (PDU) with Automatic Transfer Switch (ATS) redundancy features can be used to mitigate this weakness and enhance the uptime and robustness of single inlet devices.

Secure Racks

Intrusion and tamper proof racks or cabinets.

Secure Server Brackets

Server Bezel

Many hardware vendors distinguish themselves by having branded and styled server front-plates, combining functionality and design. Usually different modules from the same product-family share a common design language so that a filled rack (say, a combination of compute and storage nodes) has one unified design appearance. Sometimes, custom designed racks or rack doors are used, specifically in high value, cutting-edge installations or in high performance computing (HPC).

Stamping press

Stamping presses are (usually) large format hydraulic or servo driven steel/ aluminum (metal) presses where a die is used to deform, indent or cut flat blanks ("hard tool stamping", "deep drawing", "punch drawing"). The basic principle is that of hammer and anvil, with one element, the upper, called "slide" or "ram", acting as the hammer. The counter element, usually static, is the "anvil", a flat bed called the "bolster plate", and is used to fixate the "die", the form the upper ram is driven into to deform the sheet metal blank inserted in between. Stamping presses are able to deform surprisingly thick raw material, so not only sheet metal can be deformed, but also solid metal blocks can be stamepd into shape, for example to mold a frying pan.

Metal stamping press manufacturers


Sometimes inaccurately described as "metal", steel is actually an alloy. Steel is an alloy of elemental iron and carbon. Alloys are a mixture of materials, a mixture of chemical elements - and to be described as an alloy, one of the elements of this mixture has to be a metal. Alloys are formed to change the characteristics (properties) of the raw material, for example in terms of strength. Steel is harder and has improved elasticity (is less brittle) in comparison with pure iron. When an improved resistance towards oxidation (rust/ corrosion) is wanted, chromium is added to the alloy to form stainless steel. Note that the correct spelling is "steel", while spelling with an suffixed "e", as in "Steele", is an common surname ("Remington Steele", a 1980s TV series). Steel can be treated and formed in a number of ways, for example rolled, as with cold-rolled steel, or surface coated, as with galvanized steel or powder coated steel.

Steel types and grades

When speaking of steel, around the world a number of standardization schemes (grading systems) have been established to describe the different mixtures of steel (alloys). In the U.S., the standards organization "SAE International" established the "SAE designation system", sometimes also "AISI numbering" as the "American Iron and Steel Institute" (AISI) was involved in the development of the scheme. In Japan the "JIS" steel grade standards are used. In the United Kingdom the "BS" (for "British Standards" as produced by the BSI group) designation is common. In France, the "Association Française de Normalisation" (AFNOR) has set a similar numbering scheme. In Germany there are DIN steel grades, developed by the "Deutsches Institut für Normung e.V." (DIN) and a popular "Werkstoffnummer" (material no.), a 5 digit number. In Europe, generally, there are now efforts underway to replace many more European steel grade standards with a unified "EN" (European Norm) issued by the "Comité Européen de Normalisation" (CEN). While steel grade labels are nationally different, the actual mixtures are often very similar and lookup tables can be used to translate from one designation to another.

SAE steel grade numbering system

In describing the various common steel alloys, there are a number of nested "categories" of steel. The most general two types of steels are pure "Carbon steel" variants and the mixed "Alloy steels". In SAE designation Carbon steels all start with a "1". Alloy steels in SAE designation are numbered with a "2", "3", to "9" prefix, with "Manganese steels" being an exception, as it also is prefixed with a "1", from being "close" to Carbon steels.

Type SAE designation Type and composition
Carbon steels
Carbon steels 10xx Plain carbon, Mn 1.00% max
11xx Resulfurized free machining
12xx Resulfurized/rephosphorized free machining
15xx Plain carbon, Mn 1.00-1.65%
Alloy steels
Manganese steels 13xx Mn 1.75%
Nickel steels 23xx Ni 3.50%
25xx Ni 5.00%
Nickel-chromium steels 31xx Ni 1.25%, Cr 0.65-0.80%
32xx Ni 1.75%, Cr 1.07%
33xx Ni 3.50%, Cr 1.50-1.57%
34xx Ni 3.00%, Cr 0.77%
Molybdenum steels 40xx Mo 0.20-0.25%
44xx Mo 0.40-0.52%
Chromium-molybdenum steels 41xx Cr 0.50-0.95%, Mo 0.12-0.30%
Nickel-chromium-molybdenum steels 43xx Ni 1.82%, Cr 0.50-0.80%, Mo 0.25%.
47xx Ni 1.05%, Cr 0.45%, Mo 0.20-0.35%
Nickel-molybdenum steels 46xx Ni 0.85-1.82%, Mo 0.20-0.25%
48xx Ni 3.50%, Mo 0.25%
Chromium steels 50xx Cr 0.27-0.65%
51xx Cr 0.80-1.05%
50xxx Cr 0.50%, C 1.00% min
51xxx Cr 1.02%, C 1.00% min
52xxx Cr 1.45%, C 1.00% min
Chromium-vanadium steels 61xx Cr 0.60-0.95%, V 0.10-0.015%
Tungsten-chromium steels 72xx W 1.75%, Cr 0.75%
Nickel-chromium-molybdenum steels 81xx Ni 0.30%, Cr 0.40%, Mo 0.12%
86xx Ni 0.55%, Cr 0.50%, Mo 0.20%
87xx Ni 0.55%, Cr 0.50%, Mo 0.25%
88xx Ni 0.55%, Cr 0.50%, Mo 0.35%
Silicon-manganese steels 92xx Si 1.40-2.00%, Mn 0.65-0.85%, Cr 0-0.65%
Nickel-chromium-molybdenum steels 93xx Ni 3.25%, Cr 1.20%, Mo 0.12%
94xx Ni 0.45%, Cr 0.40%, Mo 0.12%
97xx Ni 0.55%, Cr 0.20%, Mo 0.20%
98xx Ni 1.00%, Cr 0.80%, Mo 0.25%

Common steel grades (types of steel)

  • DC01
    DC01 steel is a European cold-rolled steel with a yield strength of around 280 MPa and a tensile strength of in between 390-540 MPa. Other labels for DC01 are "DIN EN 10130", "St12" or "Werkstoffnummer 1.0330". It is a low-carbon steel product, coming on coils ("colts") of flat steel. DC01 is designed for cold forming applications. The "C" in its designation is for (cmp.) "cold-rolled", a manufacturing process. DC01 is a very common material for large metal stamping processes, as in producing autobody parts or in punch drawing (deep drawing). Rougly equivalent steels from other standards: SAE: SAE1008, SAE1010; DIN: FeP01, St12; JIS: SPCC; China: GB 08, GB 08F; India: O; ISO: Cr01, CR22.
  • S235JR
    S235JR, with the prefix "S" meaning "structural steel", is a European hot-rolled non-alloy (carbon) steel of moderate strength, where "235" is giving the minimum yield strength in MPa. Nominal tensile strength is in between 320-490 MPa. Other labels are "DIN EN 10025-2", "St37", "Werkstoffnummer 1.0038". The "JR" designates a specific impact resistance class. S235JR is a common raw material for cold forming applications, like in a brake press. Rougly equivalent steels from other standards: SAE: 1015, A283C (ASTM Grade C), SSGrade33; DIN: St37; JIS: SM400A, SS400; China: GB Q235A, GB Q235B, GB Q235D; India: IS226; ISO: E235B, Fe360B


Special enclosures to accept horizontally arranged subcomponents are called "Subracks" ("sub-racks"), "card cages", "Chassis cases", "rack mount chassis". Subracks employ a horizontal scheme for smaller components which are laid out horizontally inside 3U or taller 19" rack positions, defining a "smaller element standard" inside common 19-inch racks, where larger equipment is in contrary usually laid out vertically, above each other. Horizontally organized Subracks are popular in telecommunications, railway applications, industrial control cabinets and similar industrial or scientific applications. A common form-factor for Subracks is 3U and 6U height as vertically inserted "cards" usually require a certain height. Such cards, usually Eurocard (Euro board, German "Europakarte"), COMExpress cards, Single Board Computers (SBC), fiber networking cards, expansion cards for i/o or networking, are then either screwed in such Subracks or locked in place via a special clamping mechanism (for easier maintenance). Sometimes, Subracks feature connector backplanes so that inserted components (cards) with edge-connectors can act as elements of a system or computer, being connected via an internal bus, similar to ISA bus, or Europe Card Bus, S-100 bus. The unit of width measure for horizontally laid out front-plates/ components/ cards in a Subrack is "horizontal pitch", abbreviated as "HP" ("TE", "Teileinheiten" in German).

Probably the most popular layout in subracks is IEC 60297 Part 3-101 ("Subracks and associated plug-in units") or the "Eurocard format". The standard divides the net rack opening of a standard 19" rack into 84HP (84TE), where each vertical slot is 5.08mm (0.2 or 1/5 inch) wide. This layout is based on the Eurocard PCB format, measuring 100x100mm at minimum, 100x160mm (1.6mm PCB board thickness) being a common size and may increment in 60mm steps up to 400mm depth. One of these 100mm tall PCBs fits into a vertical slot of 133.35mm, adding about 30mm of "wiggle room" for mounting and connectors. Front plates are usually a little less tall, with 130mm being common. Eurocard subracks are often 6U tall while the standard defines 3U, 6U and 9U enclosures ("Baugruppenträger"). When a vertical card is double height, the 33.35mm of "wiggle room" are added once, making a double-tall Eurocard 233.35mm in height with a 267mm (10.5in) tall (266.7mm) front plate. A 9U card adds the 33.35mm height twice, making it a 366.70mm tall PCB inside a 400.50mm tall slot.

Subracks in audio

Aside from industrial applications, Subracks saw increasing popularity in audio and music production since the 1970s, where modular synthesizers, tone generators, compressors and audio fx processors are sold and assembled in 19" Subracks, in smaller modules of varying U height, but less wide.
  • Early synthesizer manufacturer Moog divided the horizontal space into units of 2.125" width, "Moog units", "MU", allowing eight modules to be placed side-by-side into a standard 19" rack opening. Later synth manufacturers, like Synthesizers.com, adopted the format for their products and established the aka label "DotCom format" for these modules. The "Moog Units" format is part of the category of "5U format" modules, all sharing a height of 5 standard rack units.
  • In the late 1990s, US synthesizer manufacturer Synthesis Technology introduced a slightly modified subrack format called "MOTM" ("Module of the Month"), where each "card" is five rack units tall (5U) and as wide as one rack unit (1U) would measure in height, 1.75" or 44.50mm, allowing 10 modules to be placed side-by-side. For these modules, it is common to have double-wide rack cabinets where 20 or 22 modules can be installed in one row, often laid out as two stacked rows, so 10U high racks for 40 or 44 modules. The "MOTM" format is part of the category of "5U format" modules, all sharing 5 standard rack units height.
  • The "500 Series" format defines "cassettes" of 3U height and 1U width. As such, it is similar to the MOTM form factor, except it is only 3U tall. The "500 Series" format is part of the category of "3U format" modules, all sharing a height of 3 standard rack units. Read the article on the (cmp.) 500 Series format for more details.
  • In Germany and the EU emerged the Eurorack format, a 3U tall format for modular synthesizer cards established by manufactuer Doepfer during the 1990s. The Eurorack specs in turn are based on IEC 60297-3 (DIN 41494) for subrack assemblies and "Eurocard" PCBs which are mounted inside. One slot in Eurorack racks is 0.20 inches wide or 5.08mm, with modules being either 2HP wide (HP for "horizontal pitch", or "TE" for German "Teileinheiten") or multiples of 2HP. So a common "slim" module is 20mm wide (20.32mm) (4HP/ 4TE) and 45mm deep, a "wide" (double wide) module is 40.3mm wide. The "Eurorack" format is part of the category of "3U format" modules, all sharing a height of 3 standard rack units.
  • T-Slot profiles

    T-Slot (sometimes "TSLOT") is a popular system of standardized aluminum construction framing, building block like profiles, that can be used to build structures, supports, cabinets, workbenches, tables and machines with a strong userbase in industrial automation and manufacturing. Aluminum profile framing is easier to cut to size, lighter and since advances in aluminum extrusion production during the 1950 and 1960s readily available. T-Slot profiles have a distinct appearance, with the eponymous channel or "slot" on one or more sides of mostly rectangular cross-section elements. This slot can be used to insert nuts ("T-Nuts") that can be fastened against the T-shaped collar of the slot to attach other elements to the profile in a simple self-centering fashion. German toys manufacturer FischerTechnik uses scaled down T-Slot building blocks with a fixed nut connector on ends, allowing blocks to be slipped in sideways, into the slots of other blocks. Similar to this building toy structure system, professional users value the easiness and swiftness that aluminum T-Slot profiles offer to build custom structures of high strength with a minimum of effort. As T-Slot has become a staple for industrial applications, many C-Parts vendors carry "slot bars" or "alu beams" under their own brand, while other vendors - like item International, 80/20 Inc., Bonnell or Misumi Group - focus on part variety, universal connectivity and consistent quality. Some smaller vendors produce slotted profile variations of T-Slot elements, aiming at specific markets, like the "maker scene". One well-know variation that has become a brandless staple of its own is V-Slot structural framing.
    T-Slot profile with T-Nut, various common T-Slot beam cross-sections and FischerTechnik bricks
    Square single T-Slot profile with three slots and a drop-in T-Nut in top position, various corss-sections of common T-Slot profiles and an assembly of Fischer Technik construction bricks

    Top-of-Rack Cabling (ToR)

    What's the meaning of "UL Listed"?

    is a safety and quality seal issued by Underwriters Laboratories, a US certification authority. There's the seal of "UL Listed" and "UL Recognized".

    Universal rails

    UPS (Uninterruptible Power Supply)

    device or system to provide backup power to connected devices in case the main power fails, due to outages or malfunctions. Some UPSs also provide power filtering and conditioning, so power sags or spikes are filtered and do no harm to valuable or sensitive devices. backup power in UPS systems is usually implemented with a backup battery, being charged during normal operation and then taking over the power draw once the main power fails fails (or is switched off or cut, as in "Maintenance Bypass" (MBP) mode). Different Types of UPS are available, offering diffeernt levels of filtering and differences in how quick power is switched from mains to backup battery power: Standby Power System (SPS), Single-Conversion Systems, Double-Conversion Systems and Multi-Mode Systems, Online, Offline and Line interactive UPSs.

    V-Slot profiles

    V-Slot (sometimes "VSLOT") is a system of slotted extruded aluminum construction framing "building block" elements that are a slightly adapted variation of T-Slot bars. V-Slot was created by OpenBuilds and is open source since a kickstarter project in 2013. With V-Slots, the slot or groove of the alu beam is shaped like a "V", with a mirrored 45 degree inward slope. This beveled "collar" of the slot improves the self-centering behavior of attached elements. Additionally, this V-shaped groove can be used to run similarly V-Shaped wheels, rollers or bearings ("V-Wheel") inside the "canal" of the aluminum profile for simple yet precise linear motion applications. V-Slot profiles are popular in DIY, the "maker scene" and are often used to build 3D printers, laser engravers or DIY CNC machines. There's a similar but unrelated system, loosely called "V Groove", of rollers and bearings with a V-Shaped groove in the middle, looking like two ball bearings fastened to each other, having a beveled bearing surface. Such V-Groove-Bearings can run on the corners of T-Slot beams or on dedicated V-Rails, offering a similar self-centering behaviour like the V-Slot system. Essentially, the idea of having 45 degree bevels and (matching) 90 degree grooves can be found in many forms and remixes in iron and aluminum construction where rectangular frictional or structural connections are made.

    Vertical Rack

    Special type of rack that allows equipment to be mounted vertically in order to save space, allow wall-mount installation in space-restrained environments or sometimes in-wall hidden installations in drywalls. Most vertical racks are designed as wall mounted racks or rack cabinets.

    Wall Mount Rack

    Rack cabinets or frames that can be bolted to a wall, usually elevated from the ground.

    Waterjet cutting

    also know as "Waterjet" is a technique where a machine uses high-pressure water to cut metal or other materials. A pump compresses water to over 30,000 psi and is then passed through a thin nozzle where a high-speed high-strength water jet stream emits. Computer controlled machines allow precise cuts similar to laser cutting. The water used may be pure water, for softer materials like plastic or foam, or abrasive water ("abrasive waterjet") where the liquid is carrying particles, like sand, to intensify the abrasiveness, the strength of the high-pressure water jet. Like laser cutting, the groove created by the jet ("kerf") is very thin. Also, waterjet has no thermal impact on the cut material and does not change the cut material's structure.

    More to read on Racks and Rack Mounting

    - 19" Rack in Wikipedia
    - Rack & Power section of the Open Compute Project
    - Data Storage Glossary

    On metal and metal forming

    The ASM Materials Education Foundation is publisher of an exhaustive series of metal handbooks for many decades now. Check the book "Sheet Metal Forming Fundamentals", edited by Taylan Altan and A. Erman Tekkaya (ISBN: 978-1-61503-842-8) and even more the "ASM Handbook Volume 14B: Metalworking: Sheet Forming" (link), edited by S.L. Semiatin (ISBN: 978-1-62708-186-3)

    Note on trademarks

    Many of the designations used by manufacturers and sellers to distinguish their products or services are claimed as trademarks. Where those designations appear in this text and Micropolis and/or the authors were aware of a trademark claim, the designations are mentioned along with their owners and may be additionally marked with a trademark symbol. Their use here in this rack mounting FAQ page is for educational use of the reader and is covered under nominative fair use. Micropolis is in no way suggesting support, sponsorship or endorsement of the owner of these trademarks. Only as much of such marks is used as is necessary to identify the trademark owner, product, or service.