I agree that the system should never be left in a unknown state boot-achive,
In the bootadm man page there is a mention of update-archive, no no logic, on
what triggers it.
" update-archive Updates current boot archive if required.
Applies to both SPARC and x86 platforms."
The end result should be, if there are no kernel changes, you should be able to
hit the power switch at any time, minus any svcs service issues,, and not see
the Solaris failsafe.
One one of the around build 49 to 51, it did seem to work, when we had some
power outages, up in Rochester NY, but now that you mention it, I had a fdisk
rounding issue on the partition on a Requested Block the on Apr. 17th that was
boot-archive the kernel could not read, here were the error messages at least
the start of them:
Apr 17 19:37:07 server1 genunix: [ID 314293 kern.info] device pciclass,[EMAIL
PROTECTED](
display#0) keeps up device [EMAIL PROTECTED],0(disk#1), but the latter is not
power managed
Apr 17 19:37:57 server1 sendmail[405]: [ID 702911 mail.alert] unable to qualify
my own domain name (server1) -- using short name
Apr 17 19:37:57 server1 sendmail[402]: [ID 702911 mail.alert] unable to qualify
my own domain name (server1) -- using short name
Apr 17 19:41:15 server1 ufs: [ID 717476 kern.notice] NOTICE: mount: not a UFS ma
gic number (0x0)
Apr 17 20:24:37 server1 genunix: [ID 773945 kern.notice] UltraDMA mode 6
selected
Apr 17 20:24:37 server1 last message repeated 1 time
Apr 17 20:24:37 server1 genunix: [ID 773945 kern.notice] UltraDMA mode 2
selected
Apr 17 20:24:37 server1 last message repeated 2 times
Apr 17 20:24:39 server1 gda: [ID 107833 kern.warning] WARNING: /[EMAIL
PROTECTED],0/pci-ide@
d/[EMAIL PROTECTED]/[EMAIL PROTECTED],0 (Disk0):
Apr 17 20:24:39 server1 Error for command 'read sector' Error Level: Fat
al
Apr 17 20:24:39 server1 gda: [ID 107833 kern.notice] Sense Key: uncorrectable
data error
Apr 17 20:24:39 server1 gda: [ID 107833 kern.notice] Vendor 'Gen-ATA ' error
code: 0x7
Apr 17 20:24:40 server1 gda: [ID 107833 kern.warning] WARNING: /[EMAIL
PROTECTED],0/pci-ide@
d/[EMAIL PROTECTED]/[EMAIL PROTECTED],0 (Disk0):
Apr 17 20:24:40 server1 Error for command 'read sector' Error Level: Fat
al
Apr 17 20:24:40 server1 gda: [ID 107833 kern.notice] Requested Block 23750522
1, Error Block: 237505248
Apr 17 20:24:40 server1 gda: [ID 107833 kern.notice] Sense Key: uncorrectable
data error
Apr 17 20:24:40 server1 gda: [ID 107833 kern.notice] Vendor 'Gen-ATA ' error
code: 0x7
Then I booted into Solaris failsafe and followed the boot-achive instructions
and ran a fsck, (though one would think you run the fsck first then the
boot-archive procedure). I shutdown again and the system came up OK, two more
times, with no Block error.
Then rebooted another time since then I had a another error when slice 4
refused to mount on /export and I tried to run Solaris fail safe and ran fsck
with a alt block and it failed, so I had to recreate the newfs and restore
over from the backup from disk1 c0d1.
booted once again last night and just a while ago. and no issues with the
boot-archive. thankfully.
The boot-archive should never be in a unknown state is, i.e unless a system
commands that call add_drv, such pkgadd or pkgrm progrgams or patches the
boot-archive should also be ran. basically when ever any of the system
utilities set the /reconfigure flag is when the update-archive should be ran
and kept in sync or a good archive state, but cut down risk of the system being
in a unstable state.
This message posted from opensolaris.org
_______________________________________________
opensolaris-discuss mailing list
[email protected]
