[LU-5484] LMV device not cleaned up after client mount failure

Related — Wed, 15 Jul 2015 14:13:37 +0000

After a master client failed to mount, the LMV device was not properly cleaned up, making it rather hard to unload all Lustre modules without rebooting the client host:

LNet: HW CPU cores: 32, npartitions: 4
alg: No test for adler32 (adler32-zlib)
alg: No test for crc32 (crc32-table)
alg: No test for crc32 (crc32-pclmul)
padlock: VIA PadLock Hash Engine not detected.
Lustre: Lustre: Build Version: jenkins-arch=x86_64,build_type=server,distro=el6,ib_stack=inkernel-25738-gc0c4f13-PRISTINE-2.6.32-431.23.3.el6
_lustre.gc0c4f13.x86_64
LNet: Added LNI 192.168.1.116@o2ib1 [8/256/0/180]
Lustre: 3633:0:(client.c:1926:ptlrpc_expire_one_request()) @@@ Request sent has timed out for slow reply: [sent 1408002952/real 1408002952]
req@ffff88082e6cfc00 x1476398102347780/t0(0) o250->MGC192.168.1.108@o2ib@192.168.1.108@o2ib:26/25 lens 400/544 e 0 to 1 dl 1408002957 ref 1 f
l Rpc:XN/0/ffffffff rc 0/-1
LustreError: 3567:0:(client.c:1083:ptlrpc_import_delay_req()) @@@ send limit expired req@ffff88082e6cf800 x1476398102347784/t0(0) o101->MGC
192.168.1.108@o2ib@192.168.1.108@o2ib:26/25 lens 328/344 e 0 to 0 dl 0 ref 2 fl Rpc:W/0/ffffffff rc 0/-1
LustreError: 3567:0:(lmv_obd.c:640:lmv_check_connect()) soaked-clilmv-ffff8808356d5000: no target configured for index 0.
LustreError: 3567:0:(llite_lib.c:274:client_common_fill_super()) cannot connect to soaked-clilmv-ffff8808356d5000: rc = -22
Lustre: Unmounted soaked-client
LustreError: 3567:0:(obd_mount.c:1342:lustre_fill_super()) Unable to mount (-22)

A subsequent "lctl dl" showed that the LMV device was still referenced. The story behind the MGS_CONNECT timeout will be covered elsewhere.

I'm working on this.

Whamcloud Community JIRA

[LU-5484] LMV device not cleaned up after client mount failure