== insanity test 3: Third Failure Mode: MDS/CLIENT Tue Sep 29 15:39:36 EDT 2026 ========================================================== 15:39:36 (1790710776)
Verify Lustre filesystem is up and running
Failing mds1 on oleg248-server
Stopping /mnt/lustre-mds1 (opts:) on oleg248-server
15:39:42 (1790710782) shut down
facet: mds1 facet_host: oleg248-server facet_failover_host: oleg248-server
Failover mds1 to oleg248-server
mount facets: mds1
Start mds1: mount -t lustre -o localrecov  lustre-mdt1/mdt1 /mnt/lustre-mds1
oleg248-server: oleg248-server.virtnet: executing set_default_debug vfstrace rpctrace dlmtrace neterror ha config ioctl super lfsck all
pdsh@oleg248-client: oleg248-server: ssh exited with exit code 1
Started lustre-MDT0000
15:40:19 (1790710819) targets are mounted
15:40:19 (1790710819) facet_failover done
oleg248-client.virtnet: executing wait_import_state_mount (FULL|IDLE) mdc.lustre-MDT0000-mdc-*.mds_server_uuid 1475 0
mdc.lustre-MDT0000-mdc-*.mds_server_uuid in FULL state after 0 sec
Test Lustre stability after MDS failover
Failing 2 CLIENTS
Request fail clients: 2, to fail: 0, failed: 0
No clients failed!
Test Lustre stability after CLIENT failure
Reintegrating CLIENTS