Pat_Barron@TRANSARC.COM (06/03/90)
A couple of weeks ago, I posted a request for information on QIC tape formats, and what's compatible with what. A number of people set me mail asking for a summary of the answers I got. This message that I recieved contains just about all the information that anyone sent me - since most people sent me these summaries of what was posted the last few times this was asked. Maybe it's time for a comp.periphs "Most Commonly Asked Questions" monthly posting... Many thanks to all who responded! --Pat. ---------- Forwarded message begins here ---------- Return-path: <analytics!rcsmith@uunet.UU.NET> X-Andrew-Authenticated-as: 0;transarc.com;Network-Mail Received: from transarc.com via trymail ID </afs/transarc.com/usr/pat/Mailbox/8aMdn9b0Bi81E4F05i>; Tue, 29 May 1990 12:01:13 -0400 (EDT) Received: from uunet.uu.net by transarc.com (5.54/3.15) id <AA01088> for Pat_Barron; Tue, 29 May 90 12:00:56 EDT Received: from analytics.UUCP by uunet.uu.net (5.61/1.14) with UUCP id AA06971; Tue, 29 May 90 12:00:46 -0400 Received: by anagld.analytics.com (5.52/smail2.5/10-10-89) id AA03681; Tue, 29 May 90 11:31:34 EDT From: rcsmith@anagld.analytics.com (Ray Smith) Message-Id: <9005291531.AA03681@anagld.analytics.com> Subject: QIC tape info To: Pat_Barron@TRANSARC.COM Date: Tue, 29 May 90 11:31:30 EDT Reply-To: rcsmith%anagld@uunet.UU.NET Return-Receipt-To: rcsmith%anagld@uunet.UU.NET X-Mailer: ELM [version 2.3test PL40] Pat, Here are two summaries of information regarding QIC tape drives. I hope they are helpful. -Ray =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Begin summary 1 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= From: Daryl Crandall <uunet!dash.mitre.org!daryl> Subject: Another summary of 1/4" tape systems Date: Mon, 30 Apr 90 10:35:30 EDT A while back I posted a request for a tutorial about cartridge tapes and then summarized the replies. I continued to get more replies on this subject and have compiled them into this summary. I'll repeat the original summary and then append the new summary. Daryl Crandall The MITRE Corporation (703) 883-7278 daryl@mitre.org ########################################################################### ########################################################################### ORIGINAL SUMMARY # Date: Mon, 9 Apr 90 11:16:31 EDT # From: Daryl Crandall <daryl@dash.mitre.org> # Message-Id: <9004093440.AA07838@dash.mitre.org> # To: sun-managers@dash.mitre.org # Subject: 1/4 inch tape summary # Status: R ########################################################################### Here is the summary to my posting about a tutorial on 1/4 inch tape drive systems. I'll repeat the original question. ########################################################################### Can somebody give me a quick lesson on the types of 1/4 inch cartridge tape drives and cartridges. I've become terminally confused about tape lengths, tape capacities, drive write densities, drive read densities, automatic density switching, compatabilities, number of tracks, QIC-II, QIC-24, QIC?? ... AAAARRRRGGGHHH! Why can't tar add to the end of a file written on a 1/4 inch tape? Why do you have to re-tension? What is the past, present, and future of 1/4 inch tape drives? Why is there air? :-) Daryl Crandall The MITRE Corporation 7525 Colshire Drive McLean, Virginia 22102 USA daryl@mitre.org (703) 883-7278 ########################################################################### Judging by the number of requests for the summary, There are apparently a lot of people who are confused on this subject. There are still several points that need to be cleared up, but I can not spend much more time on this subject. First of all there is a Sun document called: "Sun Tutorial on 1/4 Inch Tape Drives" (Sun Part No: 800-1315-05) This describes the physical tape mechanism. It does not describe the industry's evolution of different tape lengths and drive types. I won't go into the details in this document. Get it, it's worth it! Other sources of documentation: > Table 7-7 on page 84 of the System & Network Admin manual, shows > the relation between some tape aspects. > The manual pages for ST(4S) provide somewhat heavier type of info. According to what I have gleaned in the last couple of days, cartridge tape drives have either 4 or 9 tracks. I haven't gotten any information on which types of drives have 4 and which have 9 tracks but according to the Sun tutorial (above) "Tapes written on the 4-track drives can be read reliably by the 9-track drive" (re-tension if you get errors). According to the same source, tapes written on 9-track drives may (or may not) be readable on a 4-track drive. Tape re-tensioning is necessary relatively frequently to assure that the tape moves smoothly across the heads and doesn't "chatter" or jerk causing errors. Data is recorded on the tape in serpentine tracks. i.e. starting from the beginning of tape, track 1 is recorded forward until the physical end of tape is reached and the recording direction is switched to record track 2 in reverse. This process continues until all available tracks have been used. The erase head is the full width of the tape and is ONLY active during the 1st forward pass erasing ALL TRACKS on the tape. > What is the past, present, and future of 1/4 inch tape drives? > > past: QIC-11, QIC-24, 300' tape, 450' tape > present: QIC-150, 600' tape > future: QIC-525, 1000' tape > They're dead meat - but just like 2400' reels are still with us, 1/4 > inch cartridges will take awhile to decay away, reincarnating in new > and seemingly wondrous forms at times. Here is an incomplete table of densities and lengths: QIC-11 QIC-24 QIC-150 300' 32M ?? ?? 450' 45M ?? ?? 600' 65M 150M ?? Still need clarification on device assignments (/dev/st0, /dev/st8) and recording densities (QIC-11, QIC-24, QIC??) > QIC-11 and QIC-24 are just different formats for writing on the tape. > QIC-24 is newer, higher density, and has more secure head positioning > - less susceptibility to misaligned heads screwing things up. > QIC-11 and QIQ-24 write at the same density but with different > error correction schemes. One is typically rst0 and the other rst8. > The write heads don't generate fields strong enough to really > overwrite existing data; they can only write on blank tracks. > Why can't tar add to the end of a file written on a 1/4 inch tape? > You can add to a cartridge tape on which you have a tar record. > You just first use mt -f /dev/nrst8 fsf 1 (or nrst0 or whatever) > to skip the first record. The `n' refers to the non-rewinding device. > Use of /dev/rst8 would result in the tape rewinding after the skip. Editor's note: I haven't tried this but I had a couple of responses indicating that it would work. Feedback anyone? Why is there air? "We used it to make vacuum column tape drives work. Some of it got out of the labratory by mistake." "To fill my bicycle's tires." "To blow up basketballs." ########################################################################### Thanks to: dan@breeze.bellcore.com (Daniel Strick) ebersman@uunet.uu.net (Paul Ebersman) chuckles@sne42e.orl.mmc.com (chuck strickland) Len Evens <sysadm@gauss.math.nwu.edu> martin@molndal.ericsson.se stefan mochnacki <stefan@centaur.astro.utoronto.ca> bit!markm (Mark Morrissey) yamada-sun!root@nosun.West.Sun.COM (Super User) jaf@cana.Inference.Com (Jose Fernandez) ########################################################################### ########################################################################### NEW SUMMARY: I don't think you will find a 4 track cartrdige tape drive in any Sun but an obsolete Sun2. (We happen to have one!) Of course, if you have written 9 tracks on a tape, you won't be able to read 9 tracks with a 4 track drive. ########################################################################### One more thing. A 60 Meg tape on a 60 Meg drive will max out at ~47 Meg.` Try: dd if=/dev/rroot of=/dev/TAPE bs=1k and wait for the stats when it runs out. 47 Meg is what I got. ########################################################################### ----- Begin Included Message ----- > Why can't tar add to the end of a file written on a 1/4 inch tape? > You can add to a cartridge tape on which you have a tar record. > You just first use mt -f /dev/nrst8 fsf 1 (or nrst0 or whatever) > to skip the first record. The `n' refers to the non-rewinding device. > Use of /dev/rst8 would result in the tape rewinding after the skip. Editor's note: I haven't tried this but I had a couple of responses indicating that it would work. Feedback anyone? ----- End Included Message ----- As far as I have experienced this is working ok. Just use mt -f fsf and skip as many records as you want. However there is one problem coming up occationally. There was a bug in the st driver in Sun OS 4.0.0 and the bug is still there in the patched st driver for SPARCstation1. The problem is that when you add a record on a new tape, a second EOF mark is written after this one to indicate that this is the end of tape. When another record is written to the tape the second EOF mark is supposed to be overwritten. This is not always so. When you try to read the records, you always get the first one, but to get the second you have to try two times to skip the second EOF mark or you must multiply your fsf parameter with two :-). ########################################################################### Well, I considered responding to the original question, but thought that you'd probably get better info from someone else. Seems you didn't... also, I have now found a copy of the *good* info I had read awhile back on this subject. I include it below, after a couple of comments to your summary. I'd make one comment to the info I saved though, and that is that at least under 4.0 the driver *can* detect QIC-11 vs QIC-24 when reading, i.e. it doesn't matter whether you use rst0 or rst8 *when reading* (this is in fact documented in st(4S)). > Here is an incomplete table of densities and lengths: > > QIC-11 QIC-24 QIC-150 > > 300' 32M ?? ?? > > 450' 45M ?? ?? > > 600' 65M 150M ?? I do believe that the capacity difference here (as well as the comment about different densities you quote) is incorrect, i.e. the difference isn't due to QIC-11 vs QIC-24 but to 4-track vs 9-track. Since Sun more or less switched from 4-track,QIC-11 to 9-track,QIC-24, this confusion (which seems to be quite widespread) is understandable. > Still need clarification on device assignments (/dev/st0, /dev/st8) and This is well documented in ST(4S) of 4.0, I think. Hope this helps... ########################################################################### About a year ago, John Gilmore submitted to Sun-Spots a well-written description of cartridge tape formats. I've included it below. Save it and spread it around. Mike Jipping Internet: jipping@cs.hope.edu Hope College BITNET: JIPPING@HOPE Department of Computer Science Voice: Hey! ========================================================================= >Date: Thu, 18 Aug 88 02:56:03 PDT >From: hoptoad.UUCP!gnu@cgl.ucsf.edu (John Gilmore) >Subject: Cartridge tape formats and sizes (/dev/rst0 versus /dev/rst8) There seems to be massive confusion about cartridge tapes. It's really simple combinations of three different parameters. There are two variants of the mechanical tape drive -- 4-track and 9-track. The tracks are used like on an 8-track audio tape (run all the way down the tape on one track, then mechanically slide the head up or down and do another pass of the tape.) The 9-track version can store 9/4ths as much data as the 4-track version. The 9-track version will read 4-track tapes but the tapes it writes can be marginal for reading on 4-track machines, because the tracks are thinner. Mostly they work. There are two variants of the tapes themselves. One is 450 feet long (DC300XL style) and the other is 600 feet long (DC600A style). They use different magnetic coatings because the tape has to be thinner to fit 600 feet into a cartridge. Some older tape drives can only read/write the 450 foot tapes because their heads can't cope with the new magnetic coatings. There is a sense hole on the cartridge (up near the write protect tab) that lets the newer drives figure out how to set up the head for this particular tape. The actual end-of-tape sensing is done with small holes punched in the tape itself, detected with an LED, a mirror in the cartridge, and a photocell, so that works fine for either tape length. There are two variants of the bit format that controllers record on the tape. One is called QIC-11, the other is QIC-24. QIC-11 is the original Archive format (Archive Corp. started the whole 1/4" streaming cartridge business). When a standards committee got a hold of it, they changed it (of course) to QIC-24. In both cases, the tape contains 512-byte blocks of data with small headers on them. For QIC-11, the block number in the header is 8 bits; for QIC-24, the number is 24 bits. That is essentially the only difference between the two. It was changed because in unusual error recovery situations it's possible for the tape to move more than 256 blocks (at 90 inches per second and 8000 bits per inch, things go by quickly -- think about it) and the controller could lose track of where it was on the tape. Both formats hold the same amount of data on a given tape. If you make up a table of this stuff, you start seeing some familiar numbers: Tape 450' 600' Drive 4-track 20MB [no such drives used in Suns] 9-track 45MB 60MB The hardware takes care of 4-track/9-track and tape size issues, so all you have to specify in software is whether you want QIC-11 or QIC-24 formatting. /dev/rst0 is QIC-11 and /dev/rst8 is QIC-24. I linked them to /dev/rst.qic11 and /dev/rst.qic24 so I could just do it without looking it up. I believe Sun should have made the tape driver software attempt to read tapes in both QIC-11 and QIC-24 format, like reel-to-reel tape drives which will read whatever density you throw at them. They didn't, unfortunately, so if you try to read a tape that's in the other format, it looks like a totally empty tape (you get a "no data" error). Just rewind and try again with the other format. If you get "no data" in both formats, you really have a blank tape (or one recorded in yet some other random format). There are a few other manufacturers who use QIC-24 tape drives; the IBM PC/RT is one. Apollos may be another, I'm not sure. Very few of the IBM PC 1/4" tape drives use QIC-24; they all went off in different directions. I don't know of any current production machines that use QIC-11 only; it's obsolete. Sun used to make all their distribution cartridges in 4-track, QIC-11 format on 450 foot tapes [20MB], since they can be read by all Suns. Starting with SunOS 4.0, they are now making Sun-4 tapes in QIC-24 on 9-track, 600 ft tapes [60MB] which reduces the number of tapes by a factor of 3. I am not sure whether Sun-3 tapes have been switched, though I think all Sun-3's can read 60MB QIC-24 tapes unless they were upgraded from Sun-2's. However, older boot PROMs can't boot from a QIC-24 tape (they never ask the tape controller to try QIC-24 mode, and it's too dumb to do it itself), which is why you may need a boot PROM upgrade from Sun Tech Support to boot SunOS 4.0 from 1/4" tape. [I wrote the 'ar' driver for 1/4" tape on Sun-1's -- my first Unix driver, and it was really bad -- and maintained the boot code for tape drives through the first Sun-3's.] John Gilmore ########################################################################### I prepared the following fact sheet on both the 9-track tape unit which we use, and the Sun cartridges. The info on the 9-track tape drive was obtained from mtio(4), xt(4S) and a fair bit of experimentation. If you are lucky, it may also be true for your own mag tape drives. The info on the cartridge tapes comes from mtio(4), st(S) and several contributors to Sun-Spots [sorry, I've forgotten who sent what]. The bottom line is that there is very little to choose between rst0 and rst8 for the amount of data you get on the tape. However, whichever you choose, you won't be able to read the tape back on some other machines! Sun-Spots also explained the problems of booting SunOS installation tapes - If I remember correctly, the boot proms in the earlier Sun 3 workstations could only read QIC-11 format. Sun in their wisdom supplied SunOS 4 in QIC-24 format, necessitating users to upgrade their boot proms. ___ ___ ___ Andrew Yeomans PSTN: 0442 230000 ext 3371 |XXX| |XXX| | | Crosfield Electronics Ltd INTL: +44 442 230000 |XXX| |XXX| |___| Three Cherry Trees Lane Fax: 0442 232301 ___ ___ ___ Hemel Hempstead UUCP: ajy@cel.uucp |XXX| | | | | Hertfordshire or mcvax!cel!ajy@uunet.uu.net |XXX| |___| |___| HP2 7RH ___ ___ ___ England .. all opinions are my own, etc. |XXX| |XXX| | | 'The network is the bottleneck' |XXX| |XXX| |___| -------------------------- cut here ------------------------------------- Magnetic tape devices --------------------- xt - Xylogics 472 1/2 inch tape controller ------------------------------------------ Device name Maj Min Rewind on close Density Block size Fixed block size mode: /dev/mt0 8 0 Rewinding 1600 bpi 2048 bytes /dev/nmt0 =/dev/mt4 8 4 Non-rewinding 1600 bpi 2048 bytes /dev/mt8 8 8 Rewinding 6250 bpi 2048 bytes /dev/nmt8 =/dev/mt12 8 12 Non-rewinding 6250 bpi 2048 bytes Variable block size (raw) mode: /dev/rmt0 30 0 Rewinding 1600 bpi Variable /dev/nrmt0 =/dev/rmt4 30 4 Non-rewinding 1600 bpi Variable /dev/rmt8 30 8 Rewinding 6250 bpi Variable /dev/nrmt8 =/dev/rmt12 30 12 Non-rewinding 6250 bpi Variable For compatibility reasons, the non-rewinding devices can be accessed by either of the two names given. 'tar' and 'dump' default to using /dev/rmt8 (rewinding). 'mt' defaults to using /dev/rmt12 (non-rewinding). The tape density may only be changed when writing if the tape is at the load point, and if the drive is set to 'host-selectable density'. The actual recorded tape density is used when reading or when writing other than at load point, and any attempts to set it (on the front panel or by device name) are ignored. The density status lights show the selected density, not the actual recorded tape density. st - SCSI 1/4 inch cartridge tape --------------------------------- Device name Maj Min Rewind on close Format Block size /dev/rst0 18 0 Rewinding QIC-11 512*n bytes /dev/nrst0 18 4 Non-rewinding QIC-11 512*n bytes /dev/rst8 18 8 Rewinding QIC-24 512*n bytes /dev/nrst8 18 12 Non-rewinding QIC-24 512*n bytes There are two parameters to consider with 1/4 inch cartridge tapes. One is the number of tracks (4 or 9), the other is the encoding format (QIC-11 or QIC-24). QIC-11 uses 1 byte block numbers, while QIC-24 uses 3 byte block numbers. QIC-11 has fractionally greater tape density, while QIC-24 is more robust. Tapes written by a 4-track drive can be read on a 9-track drive. Tapes written by a 9-track drive may sometimes be read on a 4-track drive, but with no guarantee of success. Old Sun-2 machines can only write 4-track QIC-11 tapes. Newer Sun-2 machines can only write 9-track QIC-11 tapes. Sun-3 machines can only write 9-track QIC-11 or 9-track QIC-24 tapes. Andrew J. V. Yeomans ########################################################################### We record something like 55 Meg on 600 ft tapes regularly. The last person who claimed to have results like yours with dd discovered he had inadvertently slipped a shorter tape in. However, I do admit that wouldn't explain 47 Meg. ########################################################################### I don't know if this is a repeat msg but: The important difference in capacities seems to be related to the overhead associated with small blocking factors. Try it with bs=126 (= 63k) and things should improve. I think the question was referring to the 'r' option of tar, which, according to the man page, will "write the named files on the end of the tarfile. Note: this option *does not work* with quarter-inch archive tapes." Not that i've ever had any need to use this option. I'd just be curious to know if any of the answers you got addressed this. Just to further your confusion about tape drives (:->), I'll tell you this interesting tidbit. A while ago one of the engineers here was looking at two boxes of tape cartridges (we buy 3M carts, and they come 5 to a box). One box had DC-600As in it, the other had DC-6150s. As you probably know, the 600As are 60-megabyte types, and the 6150s are 150 megabytes. Anyhow, he noticed that each box had a `Technical Brief' written on the side; _both_ boxes briefs had this information: CHARACTERISTICS 620 ft (189.0m) 12500 ftpi 550 oersted tape ftpi stands for Flux Transitions Per Inch. If you beleive those numbers, it looks like 6150s and 600As are *exactly the same* tape! Now, I said to this engineer, "Surely 3M wouldn't sell the same tape under two model numbers!" So, I stuck a DC-600 in the drive, and gave my usual backup command to the computer -- the same command that I use with the 6150s (if you're interested, the command was `dump 5cfsu /dev/rst0 1500 /home'). The dump went without a hitch; I then did a `restore t' to see if I could read the tape, and of course, all the files were present. I didn't actually go so far as to extract the contents of the files, but having been able to extract the names makes me wonder if there is any difference at all between the 600A and the 6150. ########################################################################### More thanks to: Len Evens <sysadm@gauss.math.nwu.edu> stevej@Synopsys.COM (Steven Jukoff) Sven Ole Skrivervik <svenole@vest.sdata.no> per@erix.ericsson.se (Per Hedeland) cel!ajy@uunet.uu.net (andrew yeomans) sysadm@gauss.math.nwu.edu Tue Apr 10 11:46:28 1990 stevej@Synopsys.COM (Steven Jukoff) =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= End summary 1 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Begin summary 2 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= From: matt@iquery.UUCP (Matt Reedy) Subject: SUMMARY: cartridge tape standards Keywords: QIC cartridge tape standards Message-ID: <171@iquery.UUCP> Date: 11 Nov 89 21:15:23 GMT Organization: Programmed Intelligence, San Antonio Lines: 412 Here is the summary of the mail responses and other postings regarding my cartridge tape questions. Many, many thanks to all who contributed. matt --- Matthew Reedy UUCP: cs.utexas.edu!swrinde!dpmizar!iquery!matt Programmed Intelligence Corp. 400 N Loop 1604 E, Suite 100 San Antonio, TX 78232 (512) 490 6684 Fax: (512) 490-3590 ---------------------------------------------------------------------- Date: Thu, 26 Oct 89 18:18:11 EDT From: doc@lc15a.att.com (David W Mundhenk) This certainly is a confusing area, isn't it? I've been trying to decipher this myself, so maybe we can help each other. I have a Mountain 60MB Tape Drive on my 386 at home. It is the same one the AT&T/Olivetti machines use(d), so I'm pretty sure it will work with AT&T UNIX. I haven't tested this yet, but will, soon. I have also heard that this drive can READ tapes from AT&T 3B2 computers, I assume the SCSI ones and not the old 3B2/400 "wheezer" (floppy tape?). I will test this soon also. I have some sources written out on a 3B2/600 60MB tape I would like to have on my 386... Regarding the 'QIC' standards, there seem to be a lot of them ;-). Drives by EVEREX use QIC-02 and QIC-36. I don't know which the Mountain is, but I'll check. Anyway, that's all I have now. If you can forward any other info, or post a summary I would appreciate it! Thankx, -- Dave Mundhenk [...!att!lc15a!doc] EMAIL: ...!att!lc15a!doc | "I can't complain but | /^, VOICE: (201)-580-4943 | sometimes I still do"| / } _, , , __ #include <std.disclaimer> | - Joe Walsh | /_./ (_l |/ <~_ From: <swrinde!cs.utexas.edu!convergent.com!mitisft!dold> Date: Thu, 26 Oct 89 18:59:21 -0700 Subject: Re: Cartridge tape questions matt@iquery.UUCP (Matt Reedy): Archive makes 4 drives that we use: QIC-02 interface Full height 5945L-3 60MB SCSI interface Full height 5945S 60MB SCSI interface 1/2 height 2060S 60MB SCSI interface 1/2 height 2150S 125-150MB The Archive 2150S is the only one that I would recommend for purchase today. The Archive 2150S Viper drive can read four different formats from a pre-written tape: QIC-11 four track 20MB QIC-24 nine track 60MB QIC-120 fifteen track 125MB QIC-150 eighteen track 150MB It cannot write QIC-11 nor QIC-24. It will decide on a default write format, which could be overridden, based on the cartridge it detects, according to the spacing of the BOT hole set. DC-300 can't write the default of QIC-24. DC-600A default is QIC-120, 125MB DC-600XTD recent name change to DC600-6150 default QIC-150, 150MB. You can use DC600A with no problem, it just gives 125 instead of 150MB. DC-600XTD / DC600-6150 are still a little scarce. Looking at 600 feet of tape, laid out on a table: BeginningOfTape EndOfTape | | ________________________________________________________________________ 0 0 0 0 0 0 0 0 0 0 0 ________________________________________________________________________ ->| |<- 18" 18" ->| |<- ->| |<- 18" 18" ->| |<- ->| |<- Media Detect Distance ->| |<- DC300 - 36" DC600A - 48" DC6006150 - ?" --- Clarence A Dold - dold@tsmiti.Convergent.COM (408) 434-5293 Date: Sat, 28 Oct 89 11:08:08 MST From: swrinde!cs.utexas.edu!math.arizona.edu!rsm (Robert Maier) The following comments may prove helpful. Firstly, it is a bit incorrect to speak of a `SCSI tape drive'. A SCSI tape drive is simply a tape drive equipped with a SCSI controller, so that it can be cabled directly to a SCSI bus. The controller is frequently (though not always) built right into the drive, so it seems to be an integral part of it. Also, your question about `byte swapping' is a bit confused. It is normally the host computer and its operating system that determine the order of bytes in the data blocks on the tape. It is the tape drive itself that determines the order of bytes in the block headers, in accordance with the tape formats that it's designed to understand. But your request for a `table' is eminently reasonable. There is indeed a table with columns like tape size cartridge number number bits per (megabytes) (DC****) of tracks inch I have a copy of it in my office, but since I'm not there now I can't supply it at the moment... The difficulty in getting older tape drives to read newer tapes is simply that they have trouble resolving the tape tracks! An older tape drive finds it hard to find and read the 32 tracks per tape that the new 150MB cartridges use. In the Unix world, the tape format (i.e. the format of the headers on the tape, not the number of tracks or the number of bits per inch) is more or less standardized. That's why the above table really tells you everything that you need to know. QIC-24 format is the standard, although many drives (e.g. those on Sun workstations) will read the older, very similar QIC-11 format. (QIC-11 format was, I believe, the original quarter-inch tape format invented by Archive. There are various other QIC-* formats, but they don't show up much in the Unix world.) Appended below is a copy of an article that was posted to Usenet by John Gilmore about a year ago, that tells a lot more. Note that it was written before the new 32-track, 150MB cartridges came on the market; otherwise, it's up to date. Hope these comments help... -- Robert S. Maier | Internet: rsm@math.arizona.edu [128.196.128.99] Dept. of Math. | UUCP: ..{allegra,cmcl2,hao!noao}!arizona!amethyst!rsm Univ. of Arizona | Bitnet: maier@arizrvax Tucson, AZ 85721 | FAX: +1 602 621 8322 U.S.A. | Voice: +1 602 621 6893 / +1 602 621 2617 ---------------------------------------------------------------------- >Date: Thu, 18 Aug 88 02:56:03 PDT >From: hoptoad.UUCP!gnu@cgl.ucsf.edu (John Gilmore) >Subject: Cartridge tape formats and sizes (/dev/rst0 versus /dev/rst8) There seems to be massive confusion about cartridge tapes. It's really simple combinations of three different parameters. There are two variants of the mechanical tape drive -- 4-track and 9-track. The tracks are used like on an 8-track audio tape (run all the way down the tape on one track, then mechanically slide the head up or down and do another pass of the tape.) The 9-track version can store 9/4ths as much data as the 4-track version. The 9-track version will read 4-track tapes but the tapes it writes can be marginal for reading on 4-track machines, because the tracks are thinner. Mostly they work. There are two variants of the tapes themselves. One is 450 feet long (DC300XL style) and the other is 600 feet long (DC600A style). They use different magnetic coatings because the tape has to be thinner to fit 600 feet into a cartridge. Some older tape drives can only read/write the 450 foot tapes because their heads can't cope with the new magnetic coatings. There is a sense hole on the cartridge (up near the write protect tab) that lets the newer drives figure out how to set up the head for this particular tape. The actual end-of-tape sensing is done with small holes punched in the tape itself, detected with an LED, a mirror in the cartridge, and a photocell, so that works fine for either tape length. There are two variants of the bit format that controllers record on the tape. One is called QIC-11, the other is QIC-24. QIC-11 is the original Archive format (Archive Corp. started the whole 1/4" streaming cartridge business). When a standards committee got a hold of it, they changed it (of course) to QIC-24. In both cases, the tape contains 512-byte blocks of data with small headers on them. For QIC-11, the block number in the header is 8 bits; for QIC-24, the number is 24 bits. That is essentially the only difference between the two. It was changed because in unusual error recovery situations it's possible for the tape to move more than 256 blocks (at 90 inches per second and 8000 bits per inch, things go by quickly -- think about it) and the controller could lose track of where it was on the tape. Both formats hold the same amount of data on a given tape. If you make up a table of this stuff, you start seeing some familiar numbers: Tape 450' 600' Drive 4-track 20MB [no such drives used in Suns] 9-track 45MB 60MB The hardware takes care of 4-track/9-track and tape size issues, so all you have to specify in software is whether you want QIC-11 or QIC-24 formatting. /dev/rst0 is QIC-11 and /dev/rst8 is QIC-24. I linked them to /dev/rst.qic11 and /dev/rst.qic24 so I could just do it without looking it up. I believe Sun should have made the tape driver software attempt to read tapes in both QIC-11 and QIC-24 format, like reel-to-reel tape drives which will read whatever density you throw at them. They didn't, unfortunately, so if you try to read a tape that's in the other format, it looks like a totally empty tape (you get a "no data" error). Just rewind and try again with the other format. If you get "no data" in both formats, you really have a blank tape (or one recorded in yet some other random format). There are a few other manufacturers who use QIC-24 tape drives; the IBM PC/RT is one. Apollos may be another, I'm not sure. Very few of the IBM PC 1/4" tape drives use QIC-24; they all went off in different directions. I don't know of any current production machines that use QIC-11 only; it's obsolete. Sun used to make all their distribution cartridges in 4-track, QIC-11 format on 450 foot tapes [20MB], since they can be read by all Suns. Starting with SunOS 4.0, they are now making Sun-4 tapes in QIC-24 on 9-track, 600 ft tapes [60MB] which reduces the number of tapes by a factor of 3. I am not sure whether Sun-3 tapes have been switched, though I think all Sun-3's can read 60MB QIC-24 tapes unless they were upgraded from Sun-2's. However, older boot PROMs can't boot from a QIC-24 tape (they never ask the tape controller to try QIC-24 mode, and it's too dumb to do it itself), which is why you may need a boot PROM upgrade from Sun Tech Support to boot SunOS 4.0 from 1/4" tape. [I wrote the 'ar' driver for 1/4" tape on Sun-1's -- my first Unix driver, and it was really bad -- and maintained the boot code for tape drives through the first Sun-3's.] John Gilmore Date: 30 Oct 89 15:30:40 PST (Mon) From: swrinde!cs.utexas.edu!sactoh0!jak (Jay A. Konigsberg) I don't know how to fill in the table, but I do know that a 3B2 500/600/700 with a 60/120 MB SCSI drive can read the 60MB SCSI tapes from an AT&T 6386 running AT&T 386 Unix. Also, if the 3B2 tape was written in 60 MB format it can be read by the 6386. This is a fairly narrow case, but I believe the connection is the SCSI interface. This was done at a company I used to work for. -- ------------------------------------------------------------- Jay @ SAC-UNIX, Sacramento, Ca. UUCP=...pacbell!sactoh0!jak From: neese@adaptex.UUCP Date: 28 Oct 89 16:35:00 GMT QIC refers more to the data format than to the actual interface. With a SCSI tape drive their is also a QIC standard data format. For instance, I have a Cipher ST-150S SCSI 1/4" Tape drive that reads/writes data in QIC-120 (DC-600A) and QIC-150 (DC-600XTD, DC-6150) data formats. It will also read QIC-60 data formats (DC-600A). Whether or not the interface is a SCSI or a QIC controller, has no bearing on the data format. Hope this helps. Roy Neese Adaptec Central Field Applications Engineer UUCP @ {texbell,attctc}!cpe!adaptex!neese merch!adaptex!neese From: witold@cs.washington.edu (Witold Paluszynski) Date: 29 Oct 89 22:56:13 GMT Well, there appears to be lots of QIC standards but the answer to this question is simple: they are generally INCOMPATIBLE. I was recently shopping for a quarter inch drive with similar objectives (compatibility with other systems) and I was amazed to discover that 60 MB drives were incompatible with 150 MB drives. Yes, you can read a 60 MB tape in a 150 MB drive, but you can't write it. What I would expect is the sort of upwards compatibility as with 9 track tapes, where even the newest drives can write 800 bpi tapes. I guess the manufacturers have discovered that with that much compatibility people just keep using their old drives instead of replacing them with new ones all the time. So they were careful to prevent this kind of disaster by designing incompatible standards. I hope someone can prove me wrong. Witold What tends to happen with quarter inch tapes is that to get more data onto a tape, successive standards define more tracks on the tape. As the number of tracks on a tape increase, their width decreases (after all, its still 1/4 inch :-) When it comes to writing a track, the WIDTH of the write head which creates the track is (approx) the width of the track. Higher capacity drives therefore have narrower write heads than the older drives (QIC24 -> QIC150 -> QIC320 etc....). This means that it is impossible for a "newer" drive to write "wide tracks" (which the "older" drives can read). It IS possible however for a "newer" drive to READ tapes written with wider tracks since the read head fits "inside" tracks (as written by either drive). It is therefore possible to build a drive which reads/writes a new format but can only read older formats. It is extreemly difficult (read also: expensive) to try and build a drive to read and write all formats. ----------------------------------------------------------------- Kevin Jones. | Hewlett Packard Ltd, | Computer Peripherals Bristol, kev%hpcpbla@hplb.hpl.hp.com | Filton Road, | Stoke Gifford, Tel: 011 44 272 799910 (ext 22351) | Bristol. BS12 6QZ. | ENGLAND. ----------------------------------------------------------------- Liam R. Quin, Unixsys (UK) Ltd. lee@sq.com (a visitor to Canada, not an employee of "sq") lee@anduk.co.uk (after Christmas, when I return) As other people have pointed out, one cannot reliably write a 60M cartridge with a 150M streamer. For portability, there is no doubt that the 60M cartridges are the most useful, as this gives you compatibility with Sun, NCR, CCI, Bull, and many, many others. A QIC-24 drive can read and write QIC 11 -- the same problems exist as with 150/60M, but to a much lesser extent and it usually works. You have to tell Sun users to use /dev/rst8 (QIC 24) as it is not the default at many sites. 150 M tape drives ================= The Archive Viper is the fastest of the SCSI tape streamers I have looked at, but it is also the most expensive. The Cypher SCSI drive is acceptably fast (at least under Xenix and 386/ix), and a lot cheaper. You need 386/ix 2.0.2, I think, in order to ue the Cypher drive. This also supports the Wangtek SCSI drive, which is slower again, but not much so. The viper streams (i.e., moves continuously), even with tar (block size of 20 == 20Kbyte). With the other drives, Xenix/SCSI is generally faster, and ordinary tar seems better than gnu tar. Tweaking the SCSI disconect size may affect things. I was using the Adapttec AHA1540S SCSI host adaptor. This makes a bigger difference than the choice of tape drive, I think. I am told that the Wangetek 150M cartridge drive works with the Everex (non-SCSI) controller, which might be cheaper. I have not tried this. On some non-386 Bull systems, I have seen and used ESDI streamers (made by Cypher), but the drivers are soo pathetically written that the performance was about a factor of 60 slower than a Sun (!!!!!). I think they should go quite fast ona 386, although most ESDI controllers for the PC don't support many devices so this might not be very useful. Although they can read 60M QIC 24 (not QIC11) tapes, these drives all seem to have problems with low-density tapes, and they can neither read from nor write to them. Sometimes they simply sit there retrying the first block. 60 M tape drives ================ There are problems with these in 25MHz machines, but I have no idea why. Unixsys [sic] normally uses the Wangtek drive, although they used the Bell Technologies drive on some earlier machines. No major problems, but if you let them get too dirty, they may start running off the end of tapes. Again, xenix is *much* faster, but gnu tar is a big win under 386/ix. The Cypher 60M drive seems fine, as does the Tandberg. I have also used an Archive drive, but the extra cost might not be justified. It is worth using a Wangtek or compatible drive, as both Xenix and Unix have support for them. The Wangtek controller is a full-length board with a 25-pin connector at the back, and if you connect this to a printer you destroy the printer (or so we were told by a customer!). But we had no other real problems. Hope this is of some help. Feel free to mail me for more information. I only have names of vendors and suppliers in Britain, of course, although you are welcome to ask me for these. This posting in no way relates to official Unixsys policy. I am posting on my own behalf. Lee ---------------------------------------------------------------- -- Matthew Reedy UUCP: cs.utexas.edu!swrinde!dpmizar!iquery!matt Programmed Intelligence Corp. 400 N Loop 1604 E, Suite 100 San Antonio, TX 78232 (512) 490 6684 Fax: (512) 490-3590 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= End summary 2 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Ray Smith | UUCP: {uunet,aplcen,sundc}!anagld!rcsmith Analytics, Inc. | ARPA: rcsmith@analytics.com or Suite 200 | anagld!rcsmith@uunet.uu.net or 9891 Broken Land Parkway | RCSmith@DOCKMASTER.NCSC.MIL Columbia, MD 21046 | Voice: (301) 381-4300 Fax: (301) 381-5173 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=