Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!nrl-cmf!ames!pasteur!ucbvax!bimacs.BITNET!yedidya
From: yedidya@bimacs.BITNET (Yedidya Israel)
Newsgroups: comp.unix.wizards
Subject: cdc disk on uVAXII
Message-ID: <8805181321.AA02870@jade.berkeley.edu>
Date: 18 May 88 14:21:09 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Lines: 65

Hello wizards.

        We are running 4.3 BSD on uVAXII and installed a new cdc disk.
Its controller is simulating uda50 and identify itself as an ra80.  To
maximize disk utilization we made all partitions sit in intergral
cylinders except partitions a and b. We gave "a" the standard of 15884
and b - 33616 (not 33440) so that d will begin on a limit of a cyl.
This is the fstab entry (of course we updated /sys/vaxuba/uda.c
accordingly)

# CDC-9720-736 is an ra80 compatible.
cdc9720-736|CDC-736|CDC-9720:\
        :ty=winchester:ns#50:nt#15:nc#1633:so:\
        :pa#15884:ba#8192:fa#1024:\
        :pb#33616:bb#8192:fb#1024:\
        :pc#1224750:\
        :pd#197250:bd#4096:fd#512:\
        :pe#244500:be#8192:fe#1024:\
        :pf#244500:bf#8192:ff#1024:\
        :pg#244500:bg#8192:fg#1024:\
        :ph#244500:bh#8192:fh#1024:

We have already 2 rd53 disks + 2 rx50 disktte drives. The
configuration file looks like that

..
controller      uda0    at uba? csr 0172150             vector udintr
disk            ra0     at uda0 drive 0
disk            ra1     at uda0 drive 1
disk            ra2     at uda0 drive 2
disk            ra3     at uda0 drive 3
controller      uda1    at uba? csr 0160334             vector udintr
disk            ra4     at uda1 drive 0
..

System recognizes all devices in autoconf perfectly. But we have a few
problems.

1) swapon shouts: /dev/ra4b: device not configured
   Is there another place to tell unix about swap devices excpet in
/etc/fstab? Do unix find by itself the size of swap area (which is
not 33440 like most of disks are configured) ?

2) Error messages about ra4? looks like "ra0e: hard error sn0 OFFLINE"
                                          ^^
It is true that ra4 is the 0th disk in this controller but it should
be ra4, and the "e" always appears even if we try i/o on ra4a.

3) Sometimes even when booting successfully (successfull autocnf too)
we cannot access the disk (ra0e: hard error sn0 OFFLINE).

4) We tried to verify partitions sizes (via dd if=/dev/ra4a
of=/dev/null bs=blocksize) and all of them were o.k.  except d which
lacks two sectors, the same for all disk (c-part).

Thanks in advance for any hint.

+---------------------------------------------------------------------+
| Israel Yedidya, Math & CS Department, Bar-Ilan U, Ramat-Gan, ISRAEL |
+---------------------------------------------------------------------+
| BITNET: yedidya@bimacs                                              |
| ARPA:   yedidya%bimacs.bitnet@cunyvm.cuny.edu                       |
| CSNET:  yedidya%bimacs.bitnet%cunyvm.cuny.edu@csnet-relay           |
| UUCP:   uunet!mcvax!humus!bimacs!yedidya                            |
+---------------------------------------------------------------------+