SUMMARY: SCSI Problems on 3/260

Date: Thu Jul 02 1992 - 15:55:55 CDT


        I am sorry for the delay in posting this summary. I got replies from eight
persons. Two of the answers pointed to heat problems and one hinted that the drive
may be going bad. I got the drives, cables and the controller checked, but couldn't
find any problem with them. The problem is compounded a bit by the fact that on a
Sun 3/260 there are ESDI drives and an SCSI-ESDI Adapter. So, SCSI info alone
doesn't help much. The problem that I reported never occurred again.

Thanks to the following:

From: "Manavendra K. Thakur" <sangam!!thakur>

Sounds like you're running into heat problems. Disks will start
failing like you described if they run overheat. Check the disks -
they usualy have a filter over the exhaust fan. Perhaps the filter
has gotten clogged? Also, check the air conditioning unit and the air
flow in the machine room - make sure that cool air is actually getting
to the disks in question.

Hope this helps!
From: sangam!uunet!lsr-vax!art (Art Hays - PSTAFF)

        If you suspect cabling/termination issues, I recommend active

        1.) Use a 110ohm active terminator (this is a terminator that has a
        voltage regulator inside that adjusts the termination voltage optimally
        for cable conditions). Can be supplied by Andataco, Corey Preville,
        (301) 963-9191. They are made by a company named Methode Electronics,
        (800) 323-6858. I dont know if this company is the only supplier.
From: sangam!sunkist.West.Sun.COM!dasun!rice (Jim Rice (Rice) x6772)

The most common cause (in my experience) is cables/termination.
Check for solid connections, bent pins, pinched insulation, & overall length.

Jim Rice
From: sangam!!hammond (Kevin W. Hammond)

Could possibly be your hard drive is going bad. I've seen instances of
this on a previous VAX with a Micropolis drive that would work when freshly
powered up, but would degrade as time went on.

From: sangam!!webber (Robert D Webber)

I hope that this is more help than hindrance. I do not know the
correct diagnosis for your problem, but I thought that you might
make more headway with a copy of the SCSI standard, so I'm
enclosing a uuencoded tar file. The tar file contains compressed
components of a SCSI document: if you are thumbing through one
of those, you have my sincere apologies.

The uuencoded part should be 3826 lines long. If this document
might be usef{l to you and it does not make it intact, please
let me know and I will send you another copy'k

From: sangam!snowbird.Central.Sun.COM!ldavis!woden!gunn (David Gunn)

Do you have other devices on the SCSI bus? There is limit to its length.

You might also make sure the SCSI terminator is in place and if it is,
swap it out to make sure it's good.
From: sangam!!shankar

Check for proper termination of your SCSI chain.

Also make sure you are within the limits of reasonable
cable lengths. Typically, I have personally exceeded
the recommended cable lengths a few times and it works; but once
you have gone beyond a critical point, you start seeing such problems.

But, if you say that the machine boots up fine (absolutely without errors)
after it is powered down for some time; one obvious possibility
is that some component might be failing due to overheating;
it could be the SCSI controller OR the electronics on the drive itself.

My strategy would be:

        1) Try booting with just the root disk on the SCSI chain.

        2) If step 1. is successful, start building your SCSI chain
           with additional disks/tapes.

        3) Also make sure that your 3/260 has the correct ROM rev.
           for 4.1.1b to work.
>> This one was a typo. SunOS 4.1.1b runs only on SPARC. We run 4.1.1.
        - Bala
From: sangam!dfwdsr.SINet.SLB.COM!weissborn

check the terminator on the SCSI bus
My original posting:

> We have a Sun 3/260 with 16MB RAM, 2x327 MB Micropolis-1558 disks
> running SunOS 4.1.1b. On quite a few occasions it never comes up
> and the boot messages show the following SCSI errors. It comes up quite
> normally after a few hours if it is shutdown and brought up again.
> > SunOS Release 4.1.1 (GENERIC) #1: Sat Oct 13 06:05:48 PDT 1990
> > Copyright (c) 1983-1990, Sun Microsystems, Inc.
> > .
> > .
> > .
> > sd0 at si0 slave 0
> > sd0: <Micropolis 1558 cyl 1218 alt 2 hd 15 sec 35>
> > sd1 at si0 slave 1
> > sd1: <Micropolis 1558 cyl 1218 alt 2 hd 15 sec 35>
> > .
> > .
> > .
> > sd0: I/O request timeout
> > csr= 0x1407 bcr= 0 tcr= 0x4
> > cbsr= 0x0 (BUS FREE) cdr= 0x0 mr= 0x0 bsr= 0x0
> > target= 0, lun= 0 DMA addr= 0xef60 count= 8192 (8192)
> > cdb= 8 5 47 7a 10 0
> > last phase= 0x90 (Lost Reselect) 0 0
> > last phase= 0x90 (Lost Reselect) 4
> > last phase= 0x8c (Disconnect MSG) 0
> > last phase= 0x8 (COMMAND) 8
> > last phase= 0x81 (Arbitration) 0 0
> > last phase= 0x8d (Cmd complete MSG) 0
> > last phase= 0xc (STATUS) 0
> > last phase= 0x4 (DATA IN) 1536
> > sd1: I/O request timeout
> > csr= 0x1407 bcr= 0 tcr= 0x4
> > cbsr= 0x0 (BUS FREE) cdr= 0x0 mr= 0x0 bsr= 0x0
> > target= 0, lun= 1 DMA addr= 0x2ef60 count= 8192 (8192)
> > cdb= 8 22 93 10 10 0
> > last phase= 0x8d (Cmd complete MSG) 0
> > last phase= 0xc (STATUS) 0
> > last phase= 0x4 (DATA IN) 8192
> > last phase= 0x82 (Identify MSG) 0 0
> > last phase= 0x8b (Reselect) 0
> > last phase= 0x8c (Disconnect MSG) 0
> > last phase= 0x8 (COMMAND) 8
> > last phase= 0x81 (Arbitration) 0 0
> > .
> > .
> > .

Balasubramanian V FAX: (0212) 345149
Tata Institute of Fundamental Research Email:
Poona University Campus
Pune - 411 007. India

This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:44 CDT