[LU-4600] Test failure on test suite conf-sanity, subtest test_50h "some OSC imports are still not connected"

• Type: Bug
• Resolution: Cannot Reproduce
• Priority: Critical
• Affects Version/s: Lustre 2.7.0, Lustre 2.5.3, Lustre 2.9.0, Lustre 2.10.0
• Severity: 3
• Assignee: Alex Zhuravlev
• Reporter: Maloo

Description

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

This issue relates to the following test suite run: http://maloo.whamcloud.com/test_sets/994ad81c-8fbc-11e3-92cc-52540035b04c.

The sub-test test_50h failed with the following error:

some OSC imports are still not connected

Info required for matching: conf-sanity 50h

Issue Links

• is related to: LU-9225 conf-sanity test_50g: /usr/bin/lfs setstripe failed: error on ioctl 0x4008669a: No data available
• is related to: LU-3529 create striped directory

Comments

[Andreas Dilger]

+1 on master: https://testing.hpdd.intel.com/test_sets/657e0ec0-ee71-11e6-b34d-5254006e85c2

[nasf (Inactive)]

+1 on master:
https://testing.hpdd.intel.com/test_sets/40074146-ea7c-11e6-be3b-5254006e85c2

[Bob Glossman (Inactive)]

another on master:
https://testing.hpdd.intel.com/test_sets/d599a2c8-e2f0-11e6-84f4-5254006e85c2

[Bob Glossman (Inactive)]

another on master:
https://testing.hpdd.intel.com/test_sets/245c2112-de9e-11e6-a08b-5254006e85c2

[Bob Glossman (Inactive)]

another on master:
https://testing.hpdd.intel.com/test_sets/9fdfff5e-0751-11e6-9e5d-5254006e85c2

[James Nunez (Inactive)]

There are two different error messages found in the test_log for the failures listed in this ticket.

Some of the logs listed here have the following error in the test_log:

open(/mnt/lustre/d50h.conf-sanity/2/f50h.conf-sanity-0) error: No space left on device

which may be related to LU-7309.

Logs with this failure are at:
2015-11-22 10:46:30 - https://testing.hpdd.intel.com/test_sets/7a1ec004-9134-11e5-b507-5254006e85c2
2016-01-05 07:30:40 - https://testing.hpdd.intel.com/test_sets/33dac2da-b3aa-11e5-8114-5254006e85c2
2016-02-01 12:07:07 - https://testing.hpdd.intel.com/test_sets/9a834e80-c908-11e5-aaa9-5254006e85c2

Others test_logs mentioned in this ticket have the following error:

open(/mnt/lustre/d50h.conf-sanity/2/f50h.conf-sanity-0) error: File too large

[Niu Yawei (Inactive)]

https://testing.hpdd.intel.com/test_sets/786f02ae-6255-11e5-8cee-5254006e85c2

[Bob Glossman (Inactive)]

another seen on b2_5:
https://testing.hpdd.intel.com/test_sets/2cdb6ca2-ca74-11e4-9330-5254006e85c2

[Doug Oucharek (Inactive)]

Seen in master (2.7): https://testing.hpdd.intel.com/test_sets/218a95fe-3e0f-11e4-b06a-5254006e85c2

[Jian Yu]

Lustre Build: https://build.hpdd.intel.com/job/lustre-b2_5/84/
Distro/Arch: RHEL6.5/x86_64
Network: o2ib

https://testing.hpdd.intel.com/test_sets/9c8fcf9a-2b62-11e4-8687-5254006e85c2

Dmesg on MDS:

LustreError: 14985:0:(osp_precreate.c:719:osp_precreate_cleanup_orphans()) lustre-OST0000-osc-MDT0000: cannot cleanup orphans: rc = -5
Lustre: DEBUG MARKER: /usr/sbin/lctl conf_param lustre-OST0000.osc.active='1'
Lustre: Permanently reactivating lustre-OST0000
LustreError: 14718:0:(lod_qos.c:946:lod_alloc_specific()) can't lstripe objid [0x200000bd0:0x5:0x0]: have 0 want 1
LustreError: 167-0: lustre-OST0000-osc-MDT0000: This client was evicted by lustre-OST0000; in progress operations using this service will fail.
Lustre: DEBUG MARKER: /usr/sbin/lctl mark  conf-sanity test_50h: @@@@@@ FAIL: some OSC imports are still not connected 

[James Nunez (Inactive)]

I hit this on master (2.7): https://testing.hpdd.intel.com/test_sets/2046baf4-11e3-11e4-90ac-5254006e85c2

In the client log, I see

open(/mnt/lustre/d50h.conf-sanity/2/f50h.conf-sanity-0) error: File too large

Maybe related to LU-4340?