[sheepdog-users] Mount a snapshot

[Valerio Pachera]

2013/3/14 Liu Yuan <namei.unix at gmail.com>:
> I can't reproduce the problem. what is fs hosting /tmp/sheepfs_test/
> mountpoint?

I tried both ext3 (/tmp/sheepfs) and xfs (/mnt/sdb1).
I can tell you I was able to use it at some point.
I stopped and restart my cluster for other reasons and now I get the
problem again.
The cluster seems to be ok: status is running; vdi check works fine.

root at sheepdog001:~# collie vdi list
  Name        Id    Size    Used  Shared    Creation time   VDI id  Copies  Tag
  win2003      0   60 GB  3.4 GB  0.0 MB 2013-03-06 15:01    64c5e
3
  wheezy       0   10 GB  1.8 GB  0.0 MB 2013-03-06 15:06   9533ed
3
s squeeze      1   10 GB  1.7 GB  0.0 MB 2013-03-13 14:37   a34c67
3
s squeeze      2   10 GB  208 MB  1.5 GB 2013-03-13 14:41   a34c6b
3
s squeeze      3   10 GB  172 MB  1.6 GB 2013-03-13 14:43   a34c6c
3
  squeeze      0   10 GB  240 MB  1.5 GB 2013-03-13 14:46   a34c6d     3

root at sheepdog001:~# sheepfs /tmp/sheepfs/
failed to create sheepfs layout
root at sheepdog001:~# ls /tmp/sheepfs/
cluster
root at sheepdog001:~# ls /tmp/sheepfs/cluster/


root at sheepdog001:~# sheepfs /mnt/sdb1/sheepfs/
failed to create sheepfs layout
root at sheepdog001:~# ls /mnt/sdb1/sheepfs/
cluster

This is what I find in mount
sheepfs on /mnt/sdb1/sheepfs type fuse
(rw,nosuid,nodev,relatime,user_id=0,group_id=0,allow_other)
fusectl on /sys/fs/fuse/connections type fusectl (rw,relatime)

On sheep.log after I stopped the cluster and reboot all nodes,
re-started the cluster and
sheepfs /mnt/sdb1/sheepfs

Mar 14 16:52:45 [main] main(685) shutdown
Mar 14 16:55:49 [main] jrnl_recover(230) opening the directory
/mnt/sheepdog/journal/
Mar 14 16:55:49 [main] jrnl_recover(235) starting journal recovery
Mar 14 16:55:49 [main] jrnl_recover(291) journal recovery complete
Mar 14 16:55:49 [main] send_join_request(1060) IPv4 ip:192.168.2.41 port:7000
Mar 14 16:55:50 [main] check_host_env(377) WARN: Allowed open files
1024 too small, suggested 1024000
Mar 14 16:55:50 [main] check_host_env(386) Allowed core file size 0,
suggested unlimited
Mar 14 16:55:50 [main] main(678) sheepdog daemon (version
0.5.5_186_gf9c15bc) started
Mar 14 16:55:50 [main] update_cluster_info(844) status = 4, epoch =
25, finished: 0
Mar 14 16:55:50 [main] update_cluster_info(844) status = 1, epoch =
25, finished: 1
Mar 14 16:56:21 [io 2774] epoch_log_read(85) failed to open epoch 14 log
Mar 14 16:56:21 [io 2774] sheep_exec_req(532) failed e
Mar 14 16:56:21 [io 2774] epoch_log_read(85) failed to open epoch 13 log
Mar 14 16:56:21 [io 2774] sheep_exec_req(532) failed e
Mar 14 16:56:21 [io 2774] epoch_log_read(85) failed to open epoch 12 log
Mar 14 16:56:21 [io 2774] sheep_exec_req(532) failed e
Mar 14 16:56:21 [io 2774] sheep_exec_req(532) failed e
Mar 14 16:56:21 [io 2774] sheep_exec_req(532) failed e
Mar 14 16:56:21 [io 2774] epoch_log_read(85) failed to open epoch 11 log
Mar 14 16:56:21 [io 2774] sheep_exec_req(532) failed e
Mar 14 16:56:21 [io 2774] epoch_log_read(85) failed to open epoch 8 log
Mar 14 16:56:21 [io 2774] sheep_exec_req(532) failed e

Mount option of sdb1
/dev/sdb1 on /mnt/sdb1 type xfs (rw,relatime,attr2,delaylog,noquota)