[LU-15112] attempt to register an OST with duplicated index should fail but it does not

• Type: Bug
• Resolution: Fixed
• Priority: Major
• Fix Version/s: Lustre 2.15.0
• Severity: 3
• Assignee: Alexander Zarochentsev
• Reporter: Alexander Zarochentsev

Description

The problem is illustrated by the following conf-sanity test:

test_129()
{
        start_mds || error "MDS start failed"
        format_ost 1
        start ost1 $(ostdevname 1) $OST_MOUNT_OPTS &&
                error "start ost1 should fail" || true
        start ost1 $(ostdevname 1) $OST_MOUNT_OPTS &&
                error "second start ost1 should fail" || true
        do_facet ost1 "$TUNEFS --writeconf $(ostdevname 1)"
        start ost1 $(ostdevname 1) $OST_MOUNT_OPTS ||
                error "start ost1 failed"
        stop ost1
        stop_mds
}

OST0000 reformatted and attempted to be registered, the first attempt fails as expected with "Address already in use" error, but the second attempt succeeds:

== conf-sanity test 129: attempt to connect an OST with the same index should fail ========================================================== 17:59:44 (1634309984)
start mds service on devvm1
Starting mds1: -o localrecov  /dev/mapper/mds1_flakey /mnt/lustre-mds1
Started lustre-MDT0000
start mds service on devvm1
Starting mds2: -o localrecov  /dev/mapper/mds2_flakey /mnt/lustre-mds2
Started lustre-MDT0001
devvm1: executing wait_import_state_mount FULL mdc.lustre-MDT0000-mdc-*.mds_server_uuid
devvm1: executing wait_import_state_mount FULL mdc.lustre-MDT0001-mdc-*.mds_server_uuid
Format ost1: /dev/mapper/ost1_flakey
Starting ost1: -o localrecov  /dev/mapper/ost1_flakey /mnt/lustre-ost1
mount.lustre: mount /dev/mapper/ost1_flakey at /mnt/lustre-ost1 failed: Address already in use
The target service's index is already in use. (/dev/mapper/ost1_flakey)
Start of /dev/mapper/ost1_flakey on ost1 failed 98
Starting ost1: -o localrecov  /dev/mapper/ost1_flakey /mnt/lustre-ost1
Commit the device label on /dev/mapper/ost1_flakey
Started lustre-OST0000
 conf-sanity test_129: @@@@@@ FAIL: second start ost1 should fail 
  Trace dump:
  = ./../tests/test-framework.sh:6320:error()
  = conf-sanity.sh:9306:test_129()
  = ./../tests/test-framework.sh:6624:run_one()
  = ./../tests/test-framework.sh:6671:run_one_logged()
  = ./../tests/test-framework.sh:6497:run_test()
  = conf-sanity.sh:9313:main()
Dumping lctl log to /tmp/test_logs/1634309863/conf-sanity.test_129.*.1634310000.log
Dumping logs only on local client.
FAIL 129 (26s)

Issue Links

• is related to: LU-4966 MGS target registration should use proper UUID
• is related to: LU-17240 change test-framework to format and mount targets in parallel
• is related to: LU-15435 __req_capsule_get(): ASSERTION( msg != ((void *)0) ) failed

Comments

[Cory Spitz]

Fixed for 2.15.0.

[Gerrit Updater]

"Oleg Drokin <[e-mail]>" merged in patch https://review.whamcloud.com/45259/
Subject: LU-15112 mgc: do not ignore target registration failure
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: cefabee52586f443bfd5163f6ac0b5e1b56a9db7

[Gerrit Updater]

"Oleg Drokin <[e-mail]>" merged in patch https://review.whamcloud.com/45871/
Subject: LU-15112 ptlrpc: make rq_replied flag always correct
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 94f3f1b511609fa190cee64c7e8244f21ef70792

[Gerrit Updater]

"Alexander Zarochentsev <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/45871
Subject: LU-15112 ptlrpc: make rq_replied flag always correct
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 6e4fd9229a46fed2fbb615b040bb8017845b83bd

[Gerrit Updater]

"Alexander Zarochentsev <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/45259
Subject: LU-15112 tests: same index ost add should fail
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 18f9c246c8ec613d8aa14d0d1ca9c2d38549321f