jon@netlabs.com (Jonathan Biggar) (11/08/89)
Here is what I used on a 3/160 with a Xylogics 753 controller running
SunOS 4.0.1:
disk_type = "CDC EMD 9720-850" \
: ctlr = XD7053 \
: ncyl = 1377 : acyl = 2 : pcyl = 1379 : nhead = 15 : nsect = 67 \
: rpm = 3600 : bpt = 41088 : bps = 613
I checked on our SparcStation running 4.0.3 and found an entry identical
to what you quoted. I don't know why Sun's entry is different, but mine
works well.greg@sj.ate.slb.com (Greg Wageman) (11/09/89)
In article <2682@brazos.Rice.edu> jesse@tcgould.tn.cornell.edu (Jesse Lee) writes: >X-Sun-Spots-Digest: Volume 8, Issue 181, message 12 of 15 > >We have problem of formatting a CDC 9720-850 disk on a Sun 3/260 running >SunOS 4.0.3. We always get the error: > >Hard sectors less than nsects > >We begin to wonder if the entry of CDC 9720-850 in the file >/etc/format.dat is correct. Here is the entry found in that file: > >disk_type = "CDC 9720-850" \ > : ctlr = XD7053 \ > : ncyl = 1358 : acyl = 2 : pcyl = 1360 : nhead = 15 : nsect = 66 \ > : rpm = 3600 : bpt = 41088 : bps = 610 My CDC manual lists an 850 as having 1381 logical cylinders, 15 heads and 41,088 bytes per track. The Sun entry is going to short you some 21 cylinders of space. However, the real issue in the error message is the number of data sectors/track. This is configurable, and depends on the setting of a bank of switches on the drive, in addition to the runt-sector enable jumper (if present). You will have to consult the drive manual, as there are 3 different types of boards with different settings possible for each. Sun's entry of 66 sectors/track is a reasonable number; note however that with 610 bytes/sector (as set by the above-mentioned switches), 41,088/610 = 67 and a runt, so Sun is allowing for sector slipping. The important thing is that the drive switch configuration match the number of bytes per sector that the controller is expecting. Copyright 1989 Greg Wageman DOMAIN: greg@sj.ate.slb.com Schlumberger Technologies UUCP: {uunet,decwrl,amdahl}!sjsca4!greg San Jose, CA 95110-1397 BIX: gwage CIS: 74016,352 GEnie: G.WAGEMAN Permission granted for not-for-profit reproduction only.
shj@ames.arc.nasa.gov (Steve Jay) (11/09/89)
In v8n181, Jesse Lee (jesse@tcgould.tn.cornell.edu) asks about the format table for a CDC Sabre 850. We have a pair of Sabre 850/xy753 systems. We formatted them under OS 3.5 a long time ago, before we had the new OS 4 format program, with all the stuff in format.dat. Under OS 3, we entered the data for the drive manually, so I'm not positive of the correct entry for format.dat, but the equivalent information we used was: ncyl = 1379, acyl = 2, pcyl = 1381, nhead = 15, nsect = 67, bps = 604 I'm not sure about the entries for rpm & bpt, but I think they are correct in format.dat. The "sector" switches on the drive have to be set for 604 bytes/sector. This actually leaves room for 68 sectors/track. Formatting for 67 sectors leaves room for 1 "slip" sector per track, so minor media flaws don't require use of alternate cylinders. These parameters work out to 709 MB of space on the drive, which is further reduced to about 668 MB real space after newfs gets done with it. Note we use significantly more of the disk (1379 vs. 1358 cylinders, 67 vs. 66 sectors) than Sun's entry for the disk. I asked Sun why they appeared to purposely waste space. The answer was that when you buy a "X megabyte disk system" from Sun, you may get a CDC drive, a Fuji, or something else. The different disks have different capacities. All of the drives that might be sold as part of the "X megabyte system" are described in format.dat so that they have equivalent capacity....they purposely waste space on bigger disks to make them match the size of slightly smaller disks from another vendor. That way, if your disk breaks, the service person can replace it with a drive from another vendor, if that's what's available, and not have to worry about the customer seeing differnt capacitities. Remember, when you buy from Sun, you buy an "X megabyte system", not a specific disk drive. This seems like a sensible thing for Sun to do. However, if you buy the drive from somone other than Sun, you might as well use all of it. Anyway, we've been running with these parameters for about a year with no problems. Steve Jay Ultra Network Technologies Domain: shj@ultra.com 101 Dagget Drive Internet: ultra!shj@ames.arc.nasa.gov San Jose, CA 95134 uucp: ...ames!ultra!shj (408) 922-0100
poffen@uunet.uu.net (11/15/89)
In article <2905@brazos.Rice.edu> greg@sj.ate.slb.com (Greg Wageman) writes: >X-Sun-Spots-Digest: Volume 8, Issue 191, message 7 of 9 > >My CDC manual lists an 850 as having 1381 logical cylinders, 15 heads and >41,088 bytes per track. The Sun entry is going to short you some 21 >cylinders of space. > >However, the real issue in the error message is the number of data >sectors/track. This is configurable, and depends on the setting of a bank >of switches on the drive, in addition to the runt-sector enable jumper (if >present). You will have to consult the drive manual, as there are 3 >different types of boards with different settings possible for each. >Sun's entry of 66 sectors/track is a reasonable number; note however that >with 610 bytes/sector (as set by the above-mentioned switches), 41,088/610 >= 67 and a runt, so Sun is allowing for sector slipping. Be careful with the settings. Sun's format program REQUIRES an extra sector for sector slipping when using an XY753 (XY7053 or SMD-4). This mean that you should pick the number of sectors that best fits (more than 600, the XY753 requires 88 bytes of overhead minimum per sector, and only supports 512 byte sectors) You should set up the drive to use runt sectors, and use the round down method given in the CDC manual. Once you pick the number of sectors on the drive, tell format to use 1 LESS than this. However the bps entry should indicate the bytes/sector found in the table and the bpt entry should be the TOTAL bytes per track (including runt sectors). Here is what I use and is probably the most optimum, this uses fixed clock (not byte clock). disk_type = "CDC 9720-850" \ : ctlr = XD7053 \ : ncyl = 1379 : acyl = 2 : pcyl = 1381 : nhead = 15 : nsect = 67 \ : rpm = 3600 : bpt = 41088 : bps = 603 This means the drive must be set to 68 sectors on its switches. e-mail me for switch settings if you need it. Russ Poffenberger DOMAIN: poffen@sj.ate.slb.com Schlumberger Technologies UUCP: {uunet,decwrl,amdahl}!sjsca4!poffen 1601 Technology Drive CIS: 72401,276 San Jose, Ca. 95110 (408)437-5254