I've just noticed a huge number of errors on one of the RAID's disks. "btrfs dev stats" gives: [/dev/sdc1].write_io_errs 0 [/dev/sdc1].read_io_errs 305 [/dev/sdc1].flush_io_errs 0 [/dev/sdc1].corruption_errs 429 [/dev/sdc1].generation_errs 0 [/dev/sda1].write_io_errs 58331 [/dev/sda1].read_io_errs 57438 [/dev/sda1].flush_io_errs 37 [/dev/sda1].corruption_errs 10110 [/dev/sda1].generation_errs 0 [/dev/sdb1].write_io_errs 0 [/dev/sdb1].read_io_errs 91 [/dev/sdb1].flush_io_errs 0 [/dev/sdb1].corruption_errs 0 [/dev/sdb1].generation_errs 0 [/dev/sdd].write_io_errs 0 [/dev/sdd].read_io_errs 0 [/dev/sdd].flush_io_errs 0 [/dev/sdd].corruption_errs 0 [/dev/sdd].generation_errs 0 On the one hand, it explains the cause of problems, don't it? On another, I hardly understand how to fix such a complex problem: it seems that my stupid "check --repair" attempts corrupted FS internals heavily. Can I just remove sda from RAID-1 and run rebalance? 2017-07-27 17:02 GMT+03:00 Ivan Sizov <sivan606@xxxxxxxxx>: > My RAID-1 FS have multiple "backpointer mismatch" errors. "btrfs check > --repair" doesn't help but only increases the number of errors. > Initially, only 2 roots were affected (uncleanly deleted snapshots I > suppose). But after I ran "check --repair" new "check --readonly" > returns such errors on almost every root! > You can read a small prehistory in this mailing list thread: > https://www.mail-archive.com/linux-btrfs@xxxxxxxxxxxxxxx/msg64640.html > Recently I decided to resume my attempts to recover the FS and, at > some point I booted into my system (Fedora 25 with 4.11.3-200 kernel > with patch from the thread mentioned above). A Snapper daemon tried to > remove some corrupted snapshot and, as result, triggered btrfs-cleaner > bug. Now I can neither RW-mount the FS nor run scrub (scrub is > "aborted" in 00:00:00 after start). > > How does "--init-extent-tree" deal with backref problems? Will those > backrefs completely reconstructed? > > Can similar "backref not found" and "backpointer mismatched" errors > have very different causes and different fix scenarios? > > -- > Ivan Sizov -- Ivan Sizov -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
