Hi Gurus,
Many thanks to:
Joe Fletcher
Olle Eriksson
Joerg Bruehe
Marco Luchini
Bryan Lavelle
John J. Francini
Terry Horsnell
alan_at_nabeth
Nikola Milutinovic
for all invaluable suggestions. The problem of the disk not being recognized
was caused by the target scsi number (i.e. the actual number which is chosen
in the back of the device) I was using: 7 and 14. Our machine machine only
supports narrow scsi or 8 devices per circuit (as a matter of fact, 7 devices,
from 0 to 6, since the number 7 is reserved for the controller itself). Our
tower has already 7 devices and therefore I disconnected one of the already
present 2 Gb disks (rz11) leaving an empty slot and used its scsi target number
3. Then I could access the new disk and edit its partition table. I divided
the disk in 5 partitions, the largest ones with 4.3 Gb. But now I have a
disgusting performance problem. I was backing up with
cp -Rp * /dev/rz11d
one of the old file systems to the new disk and in the beginning everything
flows smoothly but when about 600 Mb has been done all disks hang for 90 sec.
or so; they recover only after what seems to be a general disk update happens.
The dead time increases gradually to the point it is impossible to go on.
Also, when removing large directories in the new disk the same performance
trouble happens. So I was wondering if you could have more suggestions
in order to circumvent this.
Thanks in advance,
Oyanarte Portilho
Institute of Physics
University of Brasilia, Brazil
------------
Original post:
> We bought a IBM Ultrastar 18ES 18.1Gb external disk and would like to
> install it in a DEC AXP 3000 600M box (the seller Western Scientific claims
> it fits the box), running DU 3.2. Attached to the scsi cpu connector is
> a column with six 2Gb disks and a DAT tape unit, occupying the scsi
> numbers 8, 9,..., 14. There are also two internal disks with numbers 2
> and 3; CD-ROM unit is number 4:
>
> scsi0 at tcds0 slot 0
> rz2 at scsi0 bus 0 target 2 lun 0 (DEC _ RZ28B (C) DEC 0003)
> rz3 at scsi0 bus 0 target 3 lun 0 (DEC _ RZ28B (C) DEC 0003)
> rz4 at scsi0 bus 0 target 4 lun 0 (DEC _ RRD43 (C) DEC 0064)
> scsi1 at tcds0 slot 1
> rz8 at scsi1 bus 1 target 0 lun 0 (DEC _ RZ28 (C) DEC D41C)
> rz9 at scsi1 bus 1 target 1 lun 0 (DEC _ RZ28 (C) DEC D41C)
> rz10 at scsi1 bus 1 target 2 lun 0 _(DEC RZ28 (C) DEC D41C)
> rz11 at scsi1 bus 1 target 3 lun 0 _(DEC RZ28 (C) DEC D41C)
> rz12 at scsi1 bus 1 target 4 lun 0 _(DEC RZ28 (C) DEC D41C)
> rz13 at scsi1 bus 1 target 5 lun 0 _(DEC RZ28B (C) DEC 0003)
> tz14 at scsi1 bus 1 target 6 lun 0 _(DEC TLZ06 (C)DEC 0491)
>
> We have connected the disk to the empty connector in the column with a 68
> pin cable. We tried to identify the disk as scsi number 15 and followed the
> steps:
>
> 1) cd /dev
> ./MAKEDEV rz15
>
> 2) Added the line to /etc/disktab
>
> rz30:\
> :ty=dynamic_geometry:dt=SCSI:ns#-1:nt#-1:nc#-1:
>
> 3) By giving the command
>
> disklabel -rw /dev/rrz15c rz30
>
> we get a error message:
>
> disklabel: /dev/rrz15c: No such device or address
>
> Then we tried the scsi number 7 but to no avail. We have tried those
> steps both using and not using a terminator in the external disk box.
> When turning on the station, one of the disk leds blinks; so it is
> not completely unseen. However it is not detected in the LSM prompt with
> show device or during boot. Any hints on what is wrong and how to install
> such a ultra scsi disk? Maybe the /etc/disktab entry is not appropriate.
> Suggestions?
Received on Tue Nov 30 1999 - 13:25:35 NZDT