== sanity test 160h: changelog gc thread stop upon umount, orphan records delete ========================================================== 14:39:15 (1747852755)
mdd.lustre-MDT0000.changelog_mask=+hsm
Registered 1 changelog users: 'cl9'
mdd.lustre-MDT0000.changelog_mask=+hsm
Registered 1 changelog users: 'cl9 cl10'
mdd.lustre-MDT0000.changelog_max_idle_time=10
mdd.lustre-MDT0000.changelog_gc=1
mdd.lustre-MDT0000.changelog_min_gc_interval=2
lustre-MDT0000: clear the changelog for cl9 to record #35
mds1: verifying user cl9 clear:  33 + 2 == 35
fail_loc=0x1316
total: 2 create in 0.01 seconds: 189.68 ops/second
Stopping /mnt/lustre-mds1 (opts:) on oleg220-server
fail_loc=0
Start mds1: mount -t lustre -o localrecov  /dev/mapper/mds1_flakey /mnt/lustre-mds1
oleg220-server: oleg220-server.virtnet: executing set_default_debug all all
pdsh@oleg220-client: oleg220-server: ssh exited with exit code 1
Started lustre-MDT0000
mds1: verifying first index 35 + 1 == 36
mdd.lustre-MDT0000.changelog_min_gc_interval=3600
mdd.lustre-MDT0000.changelog_gc=1
mdd.lustre-MDT0000.changelog_max_idle_time=2592000
lustre-MDT0000: changelog user 'cl10' not found
lustre-MDT0000: clear the changelog for cl9 of all records
lustre-MDT0000: Deregistered changelog user #9