Tests are done with a various lenth of instructions lists: 1, 2, 4, 8, 16, 32, 64, 128
Conclusion
All tests failed. Dues to the particular behavior of fs_stress
(random instructions sequences) a same configuration sould not
success or fail each time.
These are external view of the bugs
- sync_f function failed. One process running, sequence length
of 1 instruction
- rm fonction blocked. Try to remove an inexistant directory
and/or the directory is said to be in use.
- rm: cannot remove
`/mnt/benchs_nfs/tmp//p0/.nfs0028ac5500000002': Device or resource
busy (This file is not created by the fs_stress process).
This appened only running 4 and more fsstress process.
- fs_stress and client can not access a symlink. View from the
server, the symlink point to a non existant directory with an extra
large path length.
- Hang occurred : EIP is at nfs4_close_state.
Last operations :
nfs4_close_state
nfs4_open_revalidate
do_lookup
__link_path_walk
_spin_unlock
Fs_stress was run with 16 process and 16 operations.
- Client blocked. Fs_stress was run with 50 process and 100 operations
Client message : kernel: decode_getfattr: xdr error 10008!
Server message: kernel: svc: unknown version (0)
- Client blocked. fs_stress was run with 50 process and 100 operations
- Client messages :
VFS: Busy inodes after unmount. Self-destruct in 5 seconds. Have a nice day...
kernel: nfs4_map_errors could not handle NFSv4 error 10038
- Server message :
kernel: svc: unknown version (0)
<>
- Client blocked. fs_stress was run with 1 process and 100 operations
- Client messages :
kernel: nfs4_map_errors could not handle NFSv4 error 10038
Message repeted 10 times
kernel: decode_getfattr: xdr error 10008!
new process in ps table : nfsv4-delegretu
- Server message :
kernel: svc: unknown version (0)
Informations are given to reproduce the bugs. For now,
I don't know for each case witch function or witch functions sequence
is causing the bug.
Software configuration
- Server :
- Export options are :
/export
<world>(rw,wdelay,insecure,no_root_squash,no_subtree_check,fsid=0)
- mount options, NFSv4 :
rsize=32768,wsize=32768,timeo=14,intr,noauto
- fsstress -l 0 -d /mnt/benchs_nfs/tmp/ -n <operation_sequence_lenght> -p
<number_of_process> -r
- Linux 2.6.11-rc4-CITI_NFS4_ALL-3 patch
hardware configuration
Client
- 2 processors : Intel(R)
- Xeon(TM) CPU 2.80GHz, cache 512 KB
- Total memory: 2Gb
- Ethernet: 1Gb/s link
- Distribution : modified Fedora Core 2
Server
- 4 processors : Intel(R) Xeon(TM) CPU 2.80GHz, cache 512 KB
- Total memory: 24Gb
- Ethernet: 1Gb/s link
- Distribution : modified Fedora Core 2
