[LU-4313] error creating file during OST failover

• Type: Bug
• Resolution: Fixed
• Priority: Major
• Fix Version/s: Lustre 2.6.0, Lustre 2.4.2, Lustre 2.5.1
• Affects Version/s: Lustre 2.4.1
• Severity: 3
• Assignee: Hongchao Zhang
• Reporter: Kit Westneat (Inactive)

Description

KIT recently upgraded their test servers to Lustre 2.4.1 and are getting errors similar to LU-3645:

Nov 25 08:09:03 pfscn2 kernel: : Lustre: pfscdat2-OST0000-osc-MDT0000: Connection to pfscdat2-OST0000 (at 172.26.17.4@o2ib) was lost; in progress operations using this service will wait for recovery to complete
Nov 25 08:09:03 pfscn2 kernel: : LustreError: 12137:0:(osp_precreate.c:484:osp_precreate_send()) pfscdat2-OST0000-osc-MDT0000: can't precreate: rc = -107
Nov 25 08:10:39 pfscn2 kernel: : Lustre: pfscdat2-OST0000-osc-MDT0000: Connection restored to pfscdat2-OST0000 (at 172.26.17.3@o2ib)

What logs should we get to debug this issue? Any debug patches to apply?

Comments

[Peter Jones]

Landed for 2.4.2 and 2.6.0. This will also be included in 2.5.1

[Jian Yu]

Patch was back-ported to Lustre b2_4 branch: http://review.whamcloud.com/8468

[Kit Westneat (Inactive)]

Looks like the patch fixed the issue for the customer as well. Would it be possible to get this in 2.4.2?

Thanks,
Kit

[Kit Westneat (Inactive)]

Hi Hongchao,

Thanks for the quick response, it worked in my testing. I think the customer is going to be on vacation for a while, but I will update this ticket with their testing when they return.

Thanks,
Kit

[Hongchao Zhang]

Hi Kit,

Could you please try with http://review.whamcloud.com/#/c/8415/ ? Thanks

[Peter Jones]

Hongchao

What do you advise here?

Peter