I am having some odd ZFS performance issues and looking for some assistance on where to look to figure out what the underlying problem is.
System Config - Solaris 10 Update 3 (11/06) on a Sun v440- 118833-36 - 9-Bay JBOD with 180GB disks on two SCSI channels. 4 on one, 5 on the other..I forget the type of SCSI connection - 9-Bay is configured as a raid2z with all 9 disks, no spare On Monday, I had to swap out one of the disks in the 9-Bay. The disk wasn't bad, just had to swap it out for other reasons. I took the disk offline, pulled the disk, put in a new disk, and used the zpool replace command. 2.5 days later, the re-silver completed. During the time of the re-silver, zpool status was saying some errors had occurred but no data was lost or something. I did a zpool clear once the re-silver completed and haven't seen any errors since. (only been a .5 day though) Now when I use the storage pool, I will see the pool pause. I see this by doing a copying multiple large files at once to another disk and running zpool iostat 1 and see multiple seconds of inactivity. I also ran 'iostat -xv 1' and saw that the %w column for the disk that was replaced is at 100 all the time. On top of that, I am seeing some "scsi bus resets" messages in my /var/adm/messages log. Going back, these messages were there before I replaced the disk. Another odd thing I saw with the iostat was that most of the IO was on only one of the scsi channels, the one with 5 disks, and not much activity on the channel with 4 disks even though they are in the same raid2z pool. So...any suggestions on where to start? Not sure if I have a bad SCSI controller, bad disk, etc. Any suggestions on where to poke around would be great. Thanks! Chris This message posted from opensolaris.org _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
