[LU-8110] Seperated MGS/MDS: conf-sanity test_32a: MGS doesn't stop before test starts

• Type: Bug
• Resolution: Duplicate
• Priority: Minor
• Severity: 3
• Assignee: WC Triage
• Reporter: Maloo

Description

This issue was created by maloo for sarah_lw <[e-mail]>

This issue relates to the following test suite run: https://testing.hpdd.intel.com/test_sets/4fe49bb2-078e-11e6-b5f1-5254006e85c2.

The sub-test test_32a failed with the following error:

test_32a failed with 5

It looks like the cleanup() in conf-sanity.sh doesn't consider the separated MGS/MDS case and doesn't stop mgs

cleanup_nocli() {
        stop_ost || return 202
        stop_mds || return 201
        unload_modules_conf || return 203
}

cleanup() {
        umount_client $MOUNT || return 200
        cleanup_nocli || return $?
}

MGS console

01:38:43:LDISKFS-fs (loop0): mounted filesystem with ordered data mode. quota=off. Opts: 
01:38:43:LustreError: 15d-9: The MGS service was already started from server
01:38:43:LustreError: 19072:0:(obd_mount_server.c:1512:server_put_super()) no obd t32fs-MDT0000
01:38:43:LustreError: 19072:0:(obd_mount_server.c:140:server_deregister_mount()) t32fs-MDT0000 not registered
01:38:43:LustreError: 19072:0:(obd_mount.c:1426:lustre_fill_super()) Unable to mount  (-114)
01:38:43:Lustre: DEBUG MARKER: losetup -a
01:38:43:Lustre: DEBUG MARKER: /usr/sbin/lctl mark  conf-sanity test_32a: @@@@@@ FAIL: Mounting the MDT 
01:38:43:Lustre: DEBUG MARKER: conf-sanity test_32a: @@@@@@ FAIL: Mounting the MDT

Issue Links

• is related to: LU-8688 All Lustre test suites should run/PASS with separate MDS and MGS
• is related to: LU-11991 conf-sanity test_32xx fails for separate MGS
• is related to: LU-10717 several conf-sanity tests failed: FAIL: start mgs failed