possible smartd race condition
Gene Czarcinski
gene at czarc.net
Thu Nov 9 13:30:59 UTC 2006
I am experiencing indications by smartd of disk errors on one of my four
systems running Fedora Core 6. This has been reported and is documented in
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=214502
This problem only occurs on one system and then only running FC6 ... running
FC5 on that system has no problems.
The problem appears to be that smartd is attempting to access smart functions
on the hard disks BEFORE it has enabled the disks for smart. If I manually
enable smart with:
smartctl -s on /dev/hda
smartctl -d ata -s on /dev/sda
smartctl -d ata -s on /dev/sdb
then there are no problems.
This problem only occurs on an ASUS A8N-E motherboard with an Athlon 64 4400+
dual processor (running either x86_64 or i386 versions of FC6).
Has anyone noticed similar problems? It is difficult to get something fixed
if I am the only one seeing it (even if I can reproduce the problem).
What to look for in /var/log/messages ... something like:
hda: drive_cmd: status=0x51 { DriveReady SeekComplete Error }
hda: drive_cmd: error=0x04 { DriveStatusError }
ide: failed opcode was: 0xb0
See the bz report for more info.
--
Gene
More information about the fedora-list
mailing list