[LU-7082] conf-sanity test_90b: MDT start failed Created: 01/Sep/15 Updated: 30/Oct/15 Resolved: 07/Oct/15 |
|
| Status: | Resolved |
| Project: | Lustre |
| Component/s: | None |
| Affects Version/s: | Lustre 2.8.0 |
| Fix Version/s: | Lustre 2.8.0 |
| Type: | Bug | Priority: | Minor |
| Reporter: | Maloo | Assignee: | James Nunez (Inactive) |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | None | ||
| Issue Links: |
|
||||||||||||
| Severity: | 3 | ||||||||||||
| Rank (Obsolete): | 9223372036854775807 | ||||||||||||
| Description |
|
This issue was created by maloo for Bob Glossman <bob.glossman@intel.com> This issue relates to the following test suite run: https://testing.hpdd.intel.com/test_sets/ed8e54ea-50c6-11e5-9da9-5254006e85c2. The sub-test test_90b failed with the following error: MDT start failed this looks like Info required for matching: conf-sanity 90b |
| Comments |
| Comment by Andreas Dilger [ 02/Sep/15 ] |
|
This test was just added via http://review.whamcloud.com/14861. |
| Comment by Gregoire Pichon [ 03/Sep/15 ] |
|
Looking at MDS1 debug_log, it appears that for part 2 of test90a the fail_loc was reset to zero before the chmod commands that should consum the mod RPC slots was launch. $ grep -E "fail_loc=|cfs_fail_loc" conf_sanity_test90a_mds_debug_log.txt 00000001:02000400:0.0:1441121327.961865:0:19916:0:(debug.c:335:libcfs_debug_mark_buffer()) DEBUG MARKER: /usr/sbin/lctl set_param fail_loc=0x159 00000020:02000000:0.0:1441121328.125150:0:19480:0:(libcfs_fail.h:96:cfs_fail_check_set()) *** cfs_fail_loc=159, val=0*** 00000020:02000000:1.0:1441121328.126229:0:19887:0:(libcfs_fail.h:96:cfs_fail_check_set()) *** cfs_fail_loc=159, val=0*** 00000020:02000000:0.0:1441121328.130553:0:19480:0:(libcfs_fail.h:96:cfs_fail_check_set()) *** cfs_fail_loc=159, val=0*** 00000020:02000000:0.0:1441121328.132688:0:19480:0:(libcfs_fail.h:96:cfs_fail_check_set()) *** cfs_fail_loc=159, val=0*** 00000020:02000000:1.0:1441121328.135003:0:19887:0:(libcfs_fail.h:96:cfs_fail_check_set()) *** cfs_fail_loc=159, val=0*** 00000020:02000000:0.0:1441121328.138574:0:19480:0:(libcfs_fail.h:96:cfs_fail_check_set()) *** cfs_fail_loc=159, val=0*** 00000001:02000400:0.0:1441121328.236111:0:19971:0:(debug.c:335:libcfs_debug_mark_buffer()) DEBUG MARKER: /usr/sbin/lctl set_param fail_loc=0 00000001:02000400:1.0:1441121335.298931:0:20027:0:(debug.c:335:libcfs_debug_mark_buffer()) DEBUG MARKER: /usr/sbin/lctl set_param fail_loc=0x159 00000020:02000000:0.0:1441121335.438032:0:19478:0:(libcfs_fail.h:96:cfs_fail_check_set()) *** cfs_fail_loc=159, val=0*** 00000001:02000400:0.0:1441121335.564351:0:20082:0:(debug.c:335:libcfs_debug_mark_buffer()) DEBUG MARKER: /usr/sbin/lctl set_param fail_loc=0 I am going to add a small delay to ensure serialization occurs as expected. The MDT start failure for test_90b and other test cases, is probably due to a bad management of the error path. Will fix this also. |
| Comment by Gerrit Updater [ 03/Sep/15 ] |
|
Grégoire Pichon (gregoire.pichon@bull.net) uploaded a new patch: http://review.whamcloud.com/16215 |
| Comment by Gerrit Updater [ 07/Oct/15 ] |
|
Oleg Drokin (oleg.drokin@intel.com) merged in patch http://review.whamcloud.com/16215/ |
| Comment by Joseph Gmitter (Inactive) [ 07/Oct/15 ] |
|
Landed for 2.8 |