[LU-5246] Failure on test suite sanity test_220: error: File too large Created: 23/Jun/14 Updated: 07/Sep/16 Resolved: 10/Aug/15 |
|
| Status: | Resolved |
| Project: | Lustre |
| Component/s: | None |
| Affects Version/s: | Lustre 2.6.0 |
| Fix Version/s: | Lustre 2.8.0 |
| Type: | Bug | Priority: | Minor |
| Reporter: | Maloo | Assignee: | Emoly Liu |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | easy | ||
| Environment: |
client and server: lustre-master build #2527 |
||
| Issue Links: |
|
||||||||||||
| Severity: | 3 | ||||||||||||
| Rank (Obsolete): | 14628 | ||||||||||||
| Description |
|
This issue was created by maloo for sarah <sarah@whamcloud.com> This issue relates to the following test suite run: http://maloo.whamcloud.com/test_sets/5619217a-f9d9-11e3-9447-52540035b04c. The sub-test test_220 failed with the following error:
test log client-30vm3: Pool lustre.test_220 created CMD: client-30vm3 /usr/sbin/lctl pool_add lustre.test_220 lustre-OST0000 client-30vm3: OST lustre-OST0000_UUID added to pool lustre.test_220 preallocated objects on MDS is 45 (7841 - 7796) OST still has 0 kbytes free create 45 files @next_id... open(/mnt/lustre/d220.sanity/f37) error: File too large total: 37 creates in 0.09 seconds: 407.08 creates/second sanity test_220: @@@@@@ FAIL: test_220 failed with 3 MDS 2 console 20:56:42:Lustre: DEBUG MARKER: == sanity test 220: preallocated MDS objects still used if ENOSPC from OST == 20:56:09 (1403322969) 20:56:42:Lustre: 26010:0:(qsd_reint.c:237:qsd_reint_index()) lustre-MDT0001: II_FL_NONUNQ is set on index transfer for fid [0x200000005:0x1011:0x0], it shouldn't be 20:56:42:Lustre: 26010:0:(qsd_reint.c:237:qsd_reint_index()) Skipped 1 previous similar message 20:56:42:LustreError: 12370:0:(lod_qos.c:948:lod_alloc_specific()) can't lstripe objid [0x400002342:0x7ff:0x0]: have 0 want 1 |
| Comments |
| Comment by Jodi Levi (Inactive) [ 04/Dec/14 ] |
|
Last failure was Nov 8: https://testing.hpdd.intel.com/sub_tests/027f3980-6a1f-11e4-aeb4-5254006e85c2 |
| Comment by Andreas Dilger [ 04/Dec/14 ] |
|
It looks like there may be a defect in the test. In the DNE striped directory case, there is nothing to constrain the files being created to the same MDT as the number of precreated files that were checked at the start. I think the test needs to be fixed to create a non-striped directory on a single MDT (probably $SINGLEMDS where the precreated objects are checked). |
| Comment by Gerrit Updater [ 04/Dec/14 ] |
|
Andreas Dilger (andreas.dilger@intel.com) uploaded a new patch: http://review.whamcloud.com/12937 |
| Comment by nasf (Inactive) [ 23/Jan/15 ] |
|
Another failure instance: |
| Comment by James Nunez (Inactive) [ 25/Jan/15 ] |
|
Another failure with lustre-master tag 2.6.92 at https://testing.hpdd.intel.com/test_sets/37e63f92-9f0d-11e4-91b3-5254006e85c2 |
| Comment by Gerrit Updater [ 09/Aug/15 ] |
|
Oleg Drokin (oleg.drokin@intel.com) merged in patch http://review.whamcloud.com/12937/ |
| Comment by Peter Jones [ 10/Aug/15 ] |
|
Landed for 2.8 |