Currently disabled write-back cache on all 4 HDDs and see if it helps. We also just paniced. fsck takes about 2 hours.
If it were a 48bit LBA, or generally bad partitioning I would have expected much larger corruption, rather than a few bytes in the middle of files ..?
So far, it seems to be confined to wd2 and wd3. I am re-testing everything on wd0 and wd1 just to be sure.
Dave B wrote:
On Mon, Dec 22, 2008 at 08:02:20PM +0900, Jorgen Lundman wrote:It tests the uploaded file good, md5sums are equal. I even tried setting O_SYNC, and mode 0400 during, and after, upload, as a test. The file is correct on disk for several minutes. Even after "sync". Then suddenly it gets changed.What's the disk caching setup--per "dkctl wd0 getcache" (or whatever applies given the disk you're using: "wd1", "scsictl sd0 ...", etc.)? And, if the write-back cache was enabled, does disabling it (e.g., "dkctl wd0 setcache r") change the behavior? Cheers, --Dave
-- Jorgen Lundman | <lundman%lundman.net@localhost> Unix Administrator | +81 (0)3 -5456-2687 ext 1017 (work) Shibuya-ku, Tokyo | +81 (0)90-5578-8500 (cell) Japan | +81 (0)3 -3375-1767 (home)