[LU-428] Lustre: 16290:0:(quota_interface.c:460:quota_chk_acq_common()) still haven't managed to acquire quota space from the quota master after 10 retries (err=0, rc=0)

• Type: Bug
• Resolution: Cannot Reproduce
• Priority: Minor
• Affects Version/s: Lustre 1.8.6
• Environment: x86_64, CentOS5, 2.6.18-194.17.1.el5_lustre.1.8.5, OFED 1.5.2, 4 OSS nodes, 4 8TB OSTs/OSS, 700 clients (some o2ib, some tcp)
• Severity: 3
• Assignee: Niu Yawei (Inactive)
• Reporter: Craig Prescott

Description

We've deployed a new filesystem recently and enabled quotas. We've gotten over 1200 of these messages since we've been in production the last couple weeks:

Lustre: 16290:0:(quota_interface.c:460:quota_chk_acq_common()) still haven't managed to acquire quota space from the quota master after 10 retries (err=0, rc=0)

Some days we get none, or very few, and some days we might get 50-100. The MDS has very little load on it. We're not aware of an operational problem associated with the above messages - no one has complained to us about I/O or quota problems. But we'd like to solve whatever issue is causing these messages.

One strange thing is that when we get one of the above messages, it is always on the 10th retry, and err is always zero and rc is always zero in that case - it seems funny to me that the 10th call to acquire() is always successful even if it failed 9 times in a row prior to this.

Issue Links

• Trackbacks: Lustre 1.8.x known issues tracker While testing against Lustre b18 branch, we would hit known bugs which were already reported in Lustre Bugzilla https://bugzilla.lustre.org/. In order to move away from relying on Bugzilla, we would create a JIRA