[LU-13301] lfs setstripe to deactivated OST does not return Created: 27/Feb/20 Updated: 01/May/20 Resolved: 01/May/20 |
|
| Status: | Resolved |
| Project: | Lustre |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | Lustre 2.14.0 |
| Type: | Bug | Priority: | Minor |
| Reporter: | Emoly Liu | Assignee: | Emoly Liu |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | None | ||
| Issue Links: |
|
||||
| Severity: | 3 | ||||
| Rank (Obsolete): | 9223372036854775807 | ||||
| Description |
|
The customer found lfs setstripe to deactivated OST does not return. This issue can be reproduced by the following steps: [root@s17 ~]# lctl set_param osp.lustre-OST0001-osc-MDT0000.active=0 osp.lustre-OST0001-osc-MDT0000.active=0 [root@s20 setstripe_test]# lfs setstripe -c 1 -o 1 /mnt/lustre/tfile ...( this does not return until we re-activate the OST )... |
| Comments |
| Comment by Gerrit Updater [ 27/Feb/20 ] |
|
Emoly Liu (emoly@whamcloud.com) uploaded a new patch: https://review.whamcloud.com/37751 |
| Comment by Li Xi [ 28/Mar/20 ] |
|
Copied here for future improvement if necessary: I feel we need a more comprehensive solution for this. There are different conditions that need to be seperated. 1) OST is deactivated on purpose. In this condition, no resend/retry from client. The client should be notified that the creation failure is caused by a previous configuration command. And it might be helpful to have some help message like "To fix the failure, try command \"lctl set_param osp.lustre-OST0001-osc-MDT0000.active=0\"". No every user/administrator has that much knowledge like Mitsuhiro. 2) OST disconnects temporarily because of networks issue or restart. It seems better to cover this temporary failure and let the client application wait for the retry. 3) MDT disconnects. The current retry works well. |
| Comment by Gerrit Updater [ 01/May/20 ] |
|
Oleg Drokin (green@whamcloud.com) merged in patch https://review.whamcloud.com/37751/ |
| Comment by Peter Jones [ 01/May/20 ] |
|
Landed for 2.14 |