Messages in this thread | | | Date | Fri, 5 Nov 1999 18:16:09 +0100 | From | Jens Benecke <> | Subject | Re: Data loss on NFS mounted directory (2.2.13) |
| |
On Fri, Nov 05, 1999 at 03:41:00PM +0100, Trond Myklebust wrote:
>>> > When the download was about half finished, I 'mv'ed the file >>> > into another directory - same NFS filesystem, same partition on >>> > server, just a subdirectory. No problem, I thought, this has >>> > always worked on all machines I know. >>> Could you run a 'tcpdump -s 256' before and after the 'mv', so >>> that we can capture a write or two? I have a feeling this is a >>> server bug (BTW > > Yes. That's just about what I wanted to hear: how to trace the > > issue. (Note that I'm not calling it a bug yet...) > The dump you provided of activity after the 'mv' seemed to be missing > 'write' calls. Is this an accident, or do the 'write' really stop going > down the wire?
I copy&paste'd it from an xterm. So, the write calls you mention were not displayed by tcpdump.
> If the latter is the case, it would be useful to see what goes on exactly > at the moment around the 'mv' itself. I suspect that the server is > stating that the file handle is stale: if we can capture the first > 'write' (and the server reply) after the mv, that should tell us whether > this is indeed happening or not.
I tried this again today. With nfs_debug = 9 (as you told me). And it works perfectly. Same kernel, same machine, same setup. We have a schroedingbug here, methinks: Once you track it, it dissapears.
> Note: you might also want to look at the debugging output from 'echo "9" > > /proc/sys/sunrpc/nfs_debug' during the 'mv'.
This should be the interesting part:
nfs: write(jens/so51a_os2_49.exe(2047449815), 4096@491520) NFS: nfs_updatepage(jens/so51a_os2_49.exe 4096@491520, sync=0) NFS: find_write_request(3/2047449815, c02f5850) NFS: create_write_request(jens/so51a_os2_49.exe, 491520+4096) NFS: append_write_request(c14ac1ec, c2fcf360) NFS: 3141 schedule_write_request (async) NFS: 3140 nfs_wback_begin (jens/so51a_os2_49.exe, status=0 flags=0) NFS: 3140 nfs_wback_result (jens/so51a_os2_49.exe, status=0, flags=100) NFS: refresh_inode(3/2047449815 ct=1) NFS: remove_write_request(c14ac1ec, c2fcfb40) NFS: nfs_readdir(//jens) last message repeated 3 times NFS: refresh_inode(3/2048665601 ct=1) NFS: refresh_inode(3/2048358438 ct=1) NFS: refresh_inode(3/2047539207 ct=1) NFS: dentry_delete(Private.16/StarOffice, 0) NFS: refresh_inode(3/2047449815 ct=1) NFS: lookup(StarOffice/so51a_os2_49.exe) NFS: dentry_delete(StarOffice/so51a_os2_49.exe, 0) NFS: refresh_inode(3/2047449815 ct=1) NFS: rename(jens/so51a_os2_49.exe -> StarOffice/so51a_os2_49.exe, ct=1) NFS: 3141 nfs_wback_begin (jens/so51a_os2_49.exe, status=0 flags=0) NFS: 3141 nfs_wback_result (jens/so51a_os2_49.exe, status=0, flags=100) NFS: refresh_inode(3/2047449815 ct=1) NFS: remove_write_request(c14ac1ec, c2fcf360) NFS: dentry_delete(jens/so51a_os2_49.exe, 0) nfs: write(StarOffice/so51a_os2_49.exe(2047449815), 4096@495616) NFS: nfs_updatepage(StarOffice/so51a_os2_49.exe 4096@495616, sync=0) NFS: find_write_request(3/2047449815, c02f5378) NFS: create_write_request(StarOffice/so51a_os2_49.exe, 495616+4096) NFS: append_write_request(c14ac1ec, c2fcf360) NFS: 3152 schedule_write_request (async) nfs: write(StarOffice/so51a_os2_49.exe(2047449815), 4096@499712) NFS: nfs_updatepage(StarOffice/so51a_os2_49.exe 4096@499712, sync=0) NFS: find_write_request(3/2047449815, c0307f28) NFS: create_write_request(StarOffice/so51a_os2_49.exe, 499712+4096) NFS: append_write_request(c14ac1ec, c2fcfb40) NFS: 3153 schedule_write_request (async) NFS: 3152 nfs_wback_begin (StarOffice/so51a_os2_49.exe, status=0 flags=0) NFS: 3152 nfs_wback_result (StarOffice/so51a_os2_49.exe, status=-116, flags=100 NFS: remove_write_request(c14ac1ec, c2fcf360)
The "trouble" is, _now_ it works.
Thanks for your help anyway. =;) For next time, I know what to do. :)
-- _ciao, Jens_______________________________ http://www.pinguin.conetix.de
Windows NT indeed has very low Total Cost of Ownership. Trouble is, Microsoft _owns_ Windows NT. You just licensed it.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |