drin@nro.cs.athabascau.ca (Adrian Smith) (04/17/91)
We're having a *major* problem with a SCSI drive on a plain, 105 Meg slab. Before I get into it, I've tried NextAnswers, FAQ, my campus rep, SysAdmin, NeXT themselves (yeah, right - some help there), formatter_1.2, SCSIInquirer, TheFormatter, and various ideas which we came up with, none of which helped. So, on to the problem: We've got a Micropolis 1684-7 drive (330 Meg half-height), which we're trying to run as an external SCSI device on the slab. We've got an enclosure, a power supply, and a cable. The drive (we thought) had a problem before with its low level format, so we sent it back to Micropolis. We got it back this afternoon, and have spent the rest of the day (6 hours) trying to get it working. If we plug the cable into the external SCSI port, power up the drive, power up the slab, and then log in, we get the "Unreadable disk" requester. If we click initialize, we then get the "Are you sure" requester. If we click erase, the drive light flashes twice, then we get an "error initializing external disk:initialization error" message. We then tried using disk from a shell. It'll write the label, and it'll even read and write the drive, but if you try and init the drive, the following happens: disk> init DESTROYS ALL EXISTING DISK DATA -- really initialize? y enter host name: nova enter disk label: Mikey writing disk label Block 0 boot is "/usr/standalone/boot", ok? y creating new filesystem on /dev/rsd1a /usr/etc/newfs -n -v /dev/rsd1a /etc/mkfs /dev/rsd1a 340448 28 7 8192 1024 32 10 60 4096 t write error: 340447 wtfs: I/O error /usr/etc/newfs /dev/rsd1a failed (status 1) This wtfs error is repeatable, and is driving us nuts!! We cannot get the drive to initialize. The drive actually has: 1780 cylinders, 28 sectors/track (1024 bytes/sector - we formatted it with formatter_1.2), and 7 heads. The disktab entry written by formatter_1.2 is as follows: # MICROP 1684-07MB1036511 1684-07MB1036511|MICROP 1684-07MB1036511|MICROP 1684-07MB1036511 w/1024 byte sectors as 1 partition:\ :ty=fixed_rw_scsi:nc#1780:nt#7:ns#28:ss#1024:rm#3600:\ :fp#160:bp#0:ng#0:gs#0:ga#0:ao#0:\ :os=sdmach:z0#32:z1#96:r0=a:\ :pa#0:sa#340448:ba#8192:fa#1024:ca#32:da#4096:ra#10:oa=time:\ :ia:ta=4.3BSD: We eventually pulled the drive and plugged it onto a SmartConnex on a 386. It formatted ok, and we could then read/write the drive. Put it back on the NeXT, it dies again. To clarify what i said above, we plugged the drive in cold and tried it the recommended 2.0 way BEFORE running formatter_1.2, or any other software for that matter. The drive is brand new replacement, having been received from Micropolis today. Anybody know what a wtfs error is? And how we can fix it? We have got to the point where we think it's mkfs that is dying, but we can't figure out *why*... -drin Adrian Smith ersys!drin@nro.cs.athabascau.ca Edmonton Remote Systems: Serving Northern Alberta since 1982
louie@sayshell.umd.edu (Louis A. Mamakos) (04/18/91)
In article <5m8i11w164w@ersys.uucp> ersys!drin@nro.cs.athabascau.ca (Adrian Smith) writes: >/usr/etc/newfs -n -v /dev/rsd1a >/etc/mkfs /dev/rsd1a 340448 28 7 8192 1024 32 10 60 4096 t >write error: 340447 >wtfs: I/O error >/usr/etc/newfs /dev/rsd1a failed (status 1) This sure smells like the fstab for the disk is incorrect, and that there may not be as many blocks on the disk as you thought. You might try to "feel" around with different size to see how large the disk is. One of the things that mkfs does is diddle the last block on the disk; that's where it looks like your write error is occuring. Does the number of sectors per track include spares (that you can't put user data on)? How about spare tracks? Can you talk to the drive at all? Try something like dd if=/dev/rsd1a of=/dev/null bs=1k count=100 to try to read the first 100 blocks off the disk. louie