Hi all, I've encountered following problem: on one of my servers, kernel started to log this messages:
Dec 19 04:02:13 server kernel: vs-3050: wait_buffer_until_released: nobody releases buffer (dev 21:01, size 4096, blocknr 38174720, count 3, list 0, state 0x10019, page c156b250, (UPTODATE,CLEAN, UNLOCKED)). Still waiting (810000000) JDIRTY !JWAIT every 3 seconds. Device is IDE disk, used only for backups, so it should not be loaded much. I've observed, that it somehow happened, that backups were running twice, but both tar processes were stuck with D status (and thus unkillable). Because I was not able to cleanly umount this partition, I had to reset machine. Reiserfsck found fixable errors, log of running with --fix-fixable: Zero bit found in on-disk bitmap after the last valid bit. Fixed. vpf-10630: The on-disk and the correct bitmaps differs. Will be fixed later. There was no problem since that, but still I'm afraid it could occur again. Does anyone know, what could have caused this? Is it possible, that concurrent tar processes accessing same archive could have caused it? I doubt it, but if it is the case, it could be quite serious problem. System is based on RH7.1, kernel (almost) vanilla 2.4.27 (it has some patches applied, but none should affect VFS layer, or reiser (mostly netfilter patches, security fixes,..)) I can provide more info, if needed. Thanks much in advance. Nikola Ciprich
