This content has been marked as final. Show 10 replies
I think you need to call Sun on this one.
I think a call to Sun is in order... however I do have a long shot you could try (at your own risk or course :) )
I checked a Solaris 10 system and confirmed that stralloctty is used by specfs (sparcv9)
I then check the 32-bit version on specfs and it doesn't use stralloctty...
You could trying booting the server in 32-bit mode instead of 64-bit while this wont solve the problem is may let you boot so you can fix it....
"........To get the image to install switch the machine to 32bit mode (not supported for newer machines). To set the mode, halt the machine and at the ok prompt type setenv boot-file kernel/unix
To change the machine back to 64bit mode type the following at the ok prompt: setenv boot-file kernel/sparcv9/unix........"
NOTE: I have not ever tried this so I would test it on a machine you have locally first!!!
BTW: I don't think patching in multi user is stupid, I think if you do it (and I do) you need to quiesce the system and make sure you have a backout plan e.g. detached mirror.
I am guessing something went amiss in the patching and the kernel is stuffed i.e. half applied patch
Thanks for the tip, unfortunately the kernel/unix trick doesn't seem to work. I stil get the same error.
ok setenv boot-file kernel/unix
boot-file = kernel/unix
ok boot disk
LOM event: +1567d+3h50m1s host reset
Netra T1 200 (UltraSPARC-IIe 500MHz), No Keyboard
OpenBoot 4.0, 512 MB memory installed, Serial #xxxxxxx.
Ethernet address xxxxxxxxxxxxx, Host ID: xxxxxxxxx.
Executing last command: boot disk
Boot device: /pci@1f,0/pci@1/scsi@8/disk@0,0 File and args: kernel/unix
SunOS Release 5.10 Version Generic_127111-05 64-bit
Copyright 1983-2007 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
/kernel/fs/sparcv9/specfs: undefined symbol 'stralloctty'
WARNING: mod_load: cannot load module 'specfs'
(Can't load specfs) Program terminated
/kernel/fs/sparcv9/specfs: undefined symbol 'stralloctty'The run time linker failed to resolve the symbol 'stralloctty' when loading module 'specfs'. This is more likely due to the previous patching process that could have completed partially.
WARNING: mod_load: cannot load module 'specfs'
What you can attempt to try to do is to remove the patch:
- Boot cdrom
<ok> boot cdrom -s
- fsck the root filesystem on your root disk. (plus any other base file systems
that are mounted seperately, e.g. /opt, /usr, /var and so on.).
- mount the root filesystem(s).
- Then run patchrm(1M) on the patch ID that you have installed before the
failure. Please read through patchrm(1M) man page before enduring these
- After patch removing, umount(1M) all the file systems that you have
- Try boot from root disk now. If all goes well, you should be fine after
If you are not comfortable doing the above steps, I would suggest your best bet is to log a support call to Sun.
Hope this helps.
Sorry that I didn't read the line that you have no cdrom access :-(
Anyway, you really need something else to boot up from (either cdrom/another disk/or network) in order to fix the issue on the root disk.
Of course this is the only server that I don't have a maintenance contract for :-/
Anyway, as the only option seems to be to get the machine to boot from some other media which I'm unable to do I decided to have the machine shipped back to me. It was scheduled to be taken out of service soon anyway and it's not critical anyway (it was my secondary mailserver but I'm now supposed to use Exchange anyway, yuck).
Thanks for your help
How did you fix the problem? Did you have to rebuild? I installed 10_Recommended patch cluster and now i get the same error. I have booted with the CD-ROM -s and have mounted the root slice to /a, usr, var and opt slice's. I have then tried to remove the patch(120011-14) and have not been able to.
Any help will help...Thanks
In the end I didn't try anymore. The server was scheduled to be taken out of service anyway so I retired it a bit earlier as expected. I never bothered (=read: didn't have time) to try to get it running again as there is nothing important on that server anyway.
Maybe one of the people who responded to my request can help.
I saw this same problem on a 420R after installing patch 127111-11. I booted from the install CD-ROM in single use mode. Ran fsck on the root file system and then mounted it under /tmp/newroot. This system only has a single root partition without a separate /var, /usr, or /opt.
I then ran patchrm -R /tmp/newroot 127111-11
after that completed I was able to boot the system normally.
I had the same problem. Backing out of the patches did not help. I ended up re-installing Solaris 10 on a V490.