log.info("starting repair test type 2")
victim_osd = chooser(manager, pool, 0)
first_mon = teuthology.get_first_mon(ctx, config)
- (mon,) = ctx.cluster.only(first_mon).remotes.iterkeys()
+ (mon,) = ctx.cluster.only(first_mon).remotes.keys()
# create object
log.info("doing put and setomapval")
The config should be as follows:
- Must include the log-whitelist below
+ Must include the log-ignorelist below
Must enable filestore_debug_inject_read_err config
example:
- chef:
- install:
- ceph:
- log-whitelist:
+ log-ignorelist:
- 'candidate had a stat error'
- 'candidate had a read error'
- 'deep-scrub 0 missing, 1 inconsistent objects'
- 'scrub [0-9]+ errors'
- 'size 1 != size'
- 'attr name mismatch'
- - 'Regular scrub request, losing deep-scrub details'
+ - 'Regular scrub request, deep-scrub details will be lost'
+ - 'candidate size [0-9]+ info size [0-9]+ mismatch'
conf:
osd:
filestore debug inject read err: true
'repair_test task only accepts a dict for config'
manager = ctx.managers['ceph']
- manager.wait_for_all_up()
+ manager.wait_for_all_osds_up()
manager.raw_cluster_cmd('osd', 'set', 'noscrub')
manager.raw_cluster_cmd('osd', 'set', 'nodeep-scrub')
repair_test_2(ctx, manager, config, choose_replica)
repair_test_erasure_code(manager, hinfoerr, 'primary', "deep-scrub")
+
+ manager.raw_cluster_cmd('osd', 'unset', 'noscrub')
+ manager.raw_cluster_cmd('osd', 'unset', 'nodeep-scrub')