On Thu, 2012-03-08 at 20:10 +0100, Rafał Kupka wrote: > On Wed, 2012-03-07 at 07:30 +1100, Dave Chinner wrote: Hi, Finally I've captured this event trace. Hope it will help. > > Also, it might be worthwhile to capture an event trace (e.g. all the > > xfs_buf events) to see what occurs just before the error is > > triggered. That might tell us exactl what sequence has occurred > > leading up to the error. > > I will try to provide this information in a few days. Kernel version - Linus tree up to: 0e6e847f "xfs: stop using the page cache to back the buffer cache" It's uniprocessor virtual machine (KVM) with 384 MB RAM. Full event trace is available at https://gist.github.com/2206109 (sorry for it's *.perl extension, my mistake). Kernel log entry: [ 495.197571] XFS (vda2): I/O error occurred: meta-data dev vda2 block 0x13d80 ("xfs_trans_read_buf") error 5 buf count 8192 [ 495.198283] XFS (vda2): xfs_imap_to_bp: xfs_trans_read_buf() returned error 5. [ 501.540477] XFS (vda2): I/O error occurred: meta-data dev vda2 block 0x13d80 ("xfs_trans_read_buf") error 5 buf count 8192 [ 501.543047] XFS (vda2): xfs_imap_to_bp: xfs_trans_read_buf() returned error 5. It's possible that this "I/O error" is caused by memory pressure? Related trace line: xfsdump-2498 [000] 24195.782269: xfs_buf_ioerror: dev 254:2 bno 0x13d80 len 0x2000 hold 3 pincount 0 lock 0 error 5 flags READ|TRYLOCK|PAGES caller _xfs_buf_ioapply Events related to "bno 0x13d80": xfsdump-2498 [000] 24195.782266: xfs_buf_get: dev 254:2 bno 0x13d80 len 0x2000 hold 2 pincount 0 lock 0 flags READ|LOCK|DONT_BLOCK caller xfs_buf_read xfsdump-2498 [000] 24195.782266: xfs_buf_read: dev 254:2 bno 0x13d80 len 0x2000 hold 2 pincount 0 lock 0 flags READ|LOCK|DONT_BLOCK caller xfs_trans_read_buf xfsdump-2498 [000] 24195.782267: xfs_buf_iorequest: dev 254:2 bno 0x13d80 len 0x2000 hold 2 pincount 0 lock 0 flags READ|TRYLOCK|PAGES caller _xfs_buf_read xfsdump-2498 [000] 24195.782267: xfs_buf_hold: dev 254:2 bno 0x13d80 len 0x2000 hold 2 pincount 0 lock 0 flags READ|TRYLOCK|PAGES caller xfs_buf_iorequest xfsdump-2498 [000] 24195.782269: xfs_buf_ioerror: dev 254:2 bno 0x13d80 len 0x2000 hold 3 pincount 0 lock 0 error 5 flags READ|TRYLOCK|PAGES caller _xfs_buf_ioapply xfsdump-2498 [000] 24195.782270: xfs_buf_rele: dev 254:2 bno 0x13d80 len 0x2000 hold 3 pincount 0 lock 0 flags READ|TRYLOCK|PAGES caller xfs_buf_iorequest xfsdump-2498 [000] 24195.782985: xfs_buf_unlock: dev 254:2 bno 0x13d80 len 0x2000 hold 2 pincount 0 lock 1 flags READ|TRYLOCK|PAGES caller xfs_buf_relse xfsdump-2498 [000] 24195.782986: xfs_buf_rele: dev 254:2 bno 0x13d80 len 0x2000 hold 2 pincount 0 lock 1 flags READ|TRYLOCK|PAGES caller xfs_trans_read_buf Do this data put some light on this issue? Regards, Rafał Kupka _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs
