UFS File System Periodically Corrupted - What to check?

The system is a SunFire V440 running SunOS 5.10

About once every 3-4 weeks, the system will reboot into single user mode on its own, and then I run svcs -xv, the filesystem service and dependent services will be disabled due to the metadisk file system being corrupted.

The I've been doing is running "fsck -f -y" on /dev/md/rdsk/mdX repeatedly until the errors are cleared, and then I have backup/restore guy restore for the last known good full backup.

The metadisk consists of 2 normal internal SCSI drives which are mirrored, which don't seem to have problems based on iostat -En.

Would there be a way to troubleshoot the disks by replacing them one at a time , letting the mirror rebuild reach time, and seeing of the problem stops after one/both drives have been replaced?. I think that would be the best way to eliminate the drives as the problem.

are you patched to current? Do that before anything else. What does

fmadm faulty -ai 

show?

Replacing the drives will work if there is a hardware error. Yes.

Humm, is the reboot causing the damaged filesystem? Or is a damaged filesystem causing a reboot?

Do this occur when the mirror is split or being resilvered?