[LU-9225] conf-sanity test_50g: /usr/bin/lfs setstripe failed: error on ioctl 0x4008669a: No data available Created: 18/Mar/17 Updated: 19/Apr/17 |
|
| Status: | Open |
| Project: | Lustre |
| Component/s: | None |
| Affects Version/s: | Lustre 2.10.0 |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Minor |
| Reporter: | Maloo | Assignee: | WC Triage |
| Resolution: | Unresolved | Votes: | 0 |
| Labels: | None | ||
| Issue Links: |
|
||||||||
| Severity: | 3 | ||||||||
| Rank (Obsolete): | 9223372036854775807 | ||||||||
| Description |
|
This issue was created by maloo for Andreas Dilger <andreas.dilger@intel.com> This issue relates to the following test suite run: https://testing.hpdd.intel.com/test_sets/91648930-0a62-11e7-b5b2-5254006e85c2. The sub-test test_50g failed with the following error in the test log: osc.lustre-OST0000-osc-ffff88000005d000.ost_server_uuid in FULL state after 0 sec error on ioctl 0x4008669a for '/mnt/lustre/f50g.conf-sanity' (3): No data available error: setstripe: create file '/mnt/lustre/f50g.conf-sanity' failed: No data available conf-sanity test_50g: @@@@@@ FAIL: /usr/bin/lfs setstripe failed and the following error in the client debug log: lov_ea.c:185:lsm_unpackmd_common() lustre-clilov_UUID: OST index 1 more than OST count 1 lov_pack.c:61:lov_dump_lmm_common() objid 0x1:7025, magic 0x0bd10bd0, pattern 0x1 lov_pack.c:65:lov_dump_lmm_common() stripe_size 1048576, stripe_count 2, layout_gen 0 lov_pack.c:84:lov_dump_lmm_objects() stripe 0 idx 1 subobj 0x0:34 lov_pack.c:84:lov_dump_lmm_objects() stripe 1 idx 0 subobj 0x0:98 lov_pack.c:308:lov_unpackmd() Process leaving rc=-22 file.c:1671:ll_file_getstripe() Process entered cl_object.c:346:cl_object_getstripe() Process entered lov_object.c:1470:lov_object_getstripe() Process entered lov_object.c:1474:lov_object_getstripe() Process leaving rc=-61 cl_object.c:356:cl_object_getstripe() Process leaving rc=-61 file.c:1679:ll_file_getstripe() Process leaving rc=-61 file.c:1707:ll_lov_setstripe() Process leaving rc=-61 file.c:2384:ll_file_ioctl() Process leaving rc=-61 so it looks like OSC for OST0001 is not set up properly, even though a file is striped across that OST. I'd think that the client would at least know about the OST, even if it was inactive. The first failure that I can see in the past 4 weeks is on 2017-03-15: Info required for matching: conf-sanity 50g |
| Comments |
| Comment by Andreas Dilger [ 18/Mar/17 ] |
|
Might relate to |