[LU-5135] Test failure sanity-scrub test_5: (13) Failed to start mds

• Status: Resolved
• Project: Lustre
• Fix Version/s: Lustre 2.6.0
• Type: Bug
• Priority: Critical
• Reporter: Maloo
• Assignee: nasf (Inactive)
• Resolution: Fixed
• Severity: 3

Description

This issue was created by maloo for Nathaniel Clark <[e-mail]>

This issue relates to the following test suite run: http://maloo.whamcloud.com/test_sets/8925808c-e89e-11e3-849b-52540035b04c.

The sub-test test_5 failed with the following error:

(13) Failed to start mds4

Info required for matching: sanity-scrub 5

Comments

[Jodi Levi (Inactive)]

Fan Yong,
Can you have a look at this one and comment on your opinion of priority for 2.6?

[nasf (Inactive)]

I cannot reproduce the failure by myself, but I found some other similar failures under DNE mode. In fact, the failure is related with neither backup/restore nor LFSCK, instead, it is related with the external patch: http://review.whamcloud.com/#/c/9967/

With this patch applied, the following simple test will get faiurel in my VM environment:

test_80b()
{
        [ $MDSCOUNT -lt 2 ] &&
                skip "We need at least 2 MDSes for the test" && exit 0

        stop $SINGLEMDS || error "Fail to stop MDT0!"

        start $SINGLEMDS $(mdsdevname ${SINGLEMDS//mds/}) $MDS_MOUNT_OPTS ||
                error "Fail to start MDT0"
}
run_test 80b "verify the patch 9967"

Oleg, would you please to consider such patch? Thanks!

[Jodi Levi (Inactive)]

Patch landed to Master.

[nasf (Inactive)]

Jodi,

I mean that the patch http://review.whamcloud.com/#/c/9967/ (that has been landed to master) may cause some DNE failures. If Oleg confirm that, we need to revert such patch. So before further confirmation, we should NOT close the ticket..