I tried for each of the disks in the show-disks
setenv boot-device disk
boot -s
I kept getting the error message "Can't open boot device";
I then tried probe-scsi-all
It seems one of the disks has a problem. I thought the command will
print the boot device. It didn't. I am a newbie at this. Can someone
please give me some instructions?
I have also tried to boot from CD-ROOM and no luck.
The guy who has the root password left the company. And I am stuck
with this legacy piece of hardware.
Thanks for your help
If the system won't boot, the root password will not help. If you can
boot from CDROM, you can change the root password.
What madel "Sun Workstation" are we talking about?
What happens when you put an install CD in th drive and type:
boot -s cdrom
??
Try:
printenv boot-device
...at the OK prompt. That will show you what disk you're booting off of.
Also, if you have more than one disk in the system, it's possible that the
second one is a mirror of the OS. In that case, you can boot from that disk
instead. Probably "boot disk1" will do it, but the "showdevs" command will
confirm it.
Colin
It gave the same error "Can't open boot device"
Hi Colin
I tried
printenv boot-device
It printed a long name.
I tried booting from all the disks and got the same message.
showdevs is not working. It prints back: "showdevs ?"
The server hardware is ultra 60
Hii Richard
I did
setenv boot-device cdrom
boot -s cdrom
and it was ok
I repeated the above process for all items in the show-disks. I got
the same error message. The last disk having the value /pci@1f,4000/
ebus@1/fdthree@14,3023f0 printed the error messages
Can't read disk label
Can't open disk label package
Try typing "devalias" at the ok prompt. That is supposed to show all
current device aliases. cdrom should be one of them. Is that alias
present? Is it pointing to a CDROM drive?
What model "Sun Workstation" are we talking about? It just might
possibly be relevant!
Hi Richard
I tried to boot from cdrom. Here are the error messages:
Fatal SCSI error at script address 10 unexpected disconnect
[repeated 3 times]
Can't open boot device
The hardware is Ultra 60
Thanks
I ran test-all and here is the report
testing /pci@1f,2000/tsi, gfxp@1
testing /pci@1f,4000/scsi@2,1
testing /pci@1f,4000/scsi@2
no targets found
selftest failed. return code = -1
testing /pci@1f, 4000/scsi@3
fastal scsi error at script address 10 unexpected disconnect
[repeated 4 times]
no targets found
selftest failed, return code=-1
testing /pci@1f,4000/network@1,1
internal loopback test -- succeeded
transceiver check -- passed
testing /pci@1f,4000/ebus@1/fdthree@14,3023f0
testing floppy disk system. a formatted disk should be in the drive
no diskette, or incorrect format
selftest failed, return code=-1
Regards
I set boot-device to /pci@1f,2000/TSI,gfxp@1
and it is hung!
PLEASE HELP
It looks as if one of your disks, or your SCSI bus, has developed problems.
Boot from CDROM and see if you can mount any of your disks. It's
possible that something corrupted the boot block but that the O/S is
still intact. This can be repaired, I believe, by running installboot.
If the O/S is no longer on the disk or if it has been corrupted, you
will need to restore from backup or reinstall Solaris.
Have any changes been made to the hardware since it last worked?
Disconnecting an external disk or a tape drive could have left the SCSI
bus unterminated which would probably prevent the system from booting.
It looks like a hardware problem with your disks and/or your SCSI Host
Bus Adapter.
Hi Richard
It is hung at
Boot device : /pci@1f,2000/TSI,gfxp@1
I restarted the machine cold/hot. The same message!
I think you need professional help!!
You set the machine to boot from the graphics adapter!!!!!!
I hate to be a Nagging Nelly, but I think this is a good time to
remind everybody that you really ought to have your system's hooked up
to a UPS of some sorts. It will not only give you some number of
minutes to shut the system down in the event of a power outage, but it
also does a pretty good job of filtering the power so that there will
be no "glitches" to harm your system.
Of course, it might be considered "shutting the barn door after the
horses got loose", but hey, those aren't going to be our last horses,
are they? So put a spring-actuated door-closing mechanism on for the
next horse!
:-)
Best of luck!
what is the output of :
1) probe-scsi-all
2) probe-ide
Dean
Hi Dean
The machine is hung. There is no ok prompt.
# cp /platform/`uname -i`/ufsboot /mnt/platform/`uname -i`
for futher quries revert back
On Mar 5, 12:43 pm, "soup_or_po...@yahoo.com"
Hi Vincent
Thanks for your reply. My PROBLEM is the machine is HUNG. What I mean
is, in a wintel machine, for lack of a better example, one would go
into setup and set the boot order, so that by sticking in a floppy or
CD the boot process will go on. I know this is open to lot of
security risks. But this Sun machine is just HUNG. What are my options
besides throwing it out of the window? BTW, I operate with a limited
budget and don't have the luxury of paying for expensive tech support.
In fact, this is the last of the Sun machines in our farm. We have
moved away from Sun to Windows 2003 managed domain. Then we have a
bunch of Linux servers. One problem going against us, the room where
the Sun is housed, gets very hot due to lack of proper cooling.
However, all the problems started because of a power trip and Sun was
not hooked up to a UPS.
>On Mar 6, 10:03 pm, vincent.andu...@gmail.com wrote:
>> Hi
>> This is due UFSboot is either corrupted or deleted by mistake, so
>> restore it and reboot from CD-ROM
>Thanks for your reply. My PROBLEM is the machine is HUNG. What I mean
>is, in a wintel machine, for lack of a better example, one would go
>into setup and set the boot order, so that by sticking in a floppy or
>CD the boot process will go on. I know this is open to lot of
>security risks. But this Sun machine is just HUNG.
You need to get to the OBP (roughly equivalent to the BIOS). If you
have a serial console, send a break signal. If you are using remote
management such as ALOM, type a `reset' command and then reconnect
to the console. If you are using a Sun keyboard, type Stop-A. In
any case, you should see the `ok' prompt from the OBP. You may need
to power-cycle the server before interrupting it. Once you have the
OBP prompt, start with something like this:
setenv auto-boot? false
reset-all
probe-scsi-all
--
-Gary Mills- -Unix Support- -U of M Academic Computing and Networking-
Know any sailors in need of a boat anchor?
Sorry, couldn't resist! <:-)
> The server hardware is ultra 60
Consider purchasing a used ultra 60 replacement. Several used Sun
stores have them available for a few hundred dollars. I've purchased
equipment from solarsystems.com, and they price one at $550, fully
warranted, or you can take your chances on Ebay for probably half to
2/3rds that price. (Look for a company with a DOA or other warranty
if you need safety.)
Hook up the old hard drive to the "new" machine and see if it
recognizes it. If possible, mount it, and/or "fsck" it.
Depending on whether the problem is in the machine or on the disk, you
might be able to recover the data. And you won't need the old root
password.
On Fri, 8 Mar 2007, soup_o...@yahoo.com wrote:
> Hi Vincent
> Thanks for your reply. My PROBLEM is the machine is HUNG. What I mean
> is, in a wintel machine, for lack of a better example, one would go
> into setup and set the boot order, so that by sticking in a floppy or
> CD the boot process will go on. I know this is open to lot of
> security risks. But this Sun machine is just HUNG. What are my options
> besides throwing it out of the window? BTW, I operate with a limited
> budget and don't have the luxury of paying for expensive tech support.
> In fact, this is the last of the Sun machines in our farm. We have
> moved away from Sun to Windows 2003 managed domain. Then we have a
> bunch of Linux servers. One problem going against us, the room where
> the Sun is housed, gets very hot due to lack of proper cooling.
> However, all the problems started because of a power trip and Sun was
> not hooked up to a UPS.
When you say HUNG, this doesnt work for me in the context of the header
about not finding the boot-device...
if its searching for a boot-device it is NOT hung, it has a broken or
corrupt boot-device
setenv auto-boot? false
reset-all
probe-scsi-all
what does that show?
Do you have a CD-Player that works in it?
If so, popin a solaris CD (any version really, but one at same or later
release as the OS was if preferable)
boot cdrom-s
use format to see if you have a responding drive or not, unless that
is evident from the earlier probe-scsi-all command
if it does respond, check its partitions layout and try to mount them
one at a time, if they kick of errors, try fsck on them and see if
that works...
usualy the [br]oot is on slice 0 on the disk, but i've seen people put
it in other places for reasons unknown...
if all checks out and it still wont boot... then you need to install a
new bootblock on it
in your case that would probably mean...
/usr/sbin/installboot /usr/platform/sun4u/lib/fs/ufs /dev/rdsk/c0t0d0s0
but if its another bootplaceyou will have to modify that...
if it still doesnt work, it looks like your boot-disk is truly corrupt
but if it was it would probably have shown much earlier...
you might also check for a possible root-mirror, and try the same
procedure on that one if there is one... since you tried to boot every
single disk I doubt there is one... but its worth looking into...
just try...
reset
boot disk0
if it fails
reset
boot disk1
...
if all fails and no boot but the disks responds and look ok, you should
probably use the backup and restore it... thats whats it for anyway...
if no disks respond... then get a replacement disk, and use the backup or
reinstall if you dont have a backup...
if someone complains about lost data or long recover times, they should
have had it at least backed up and prefereably mirrored... if they dont,
its realy not your fault...
/Johan A
Kindly accept my best regards
Thank you, Thank you, Thank you
Thank you a million times.