[LU-4480] cfs_fail_timeout id 214 sleeping for -1000ms

• Type: Bug
• Resolution: Fixed
• Priority: Critical
• Fix Version/s: Lustre 2.7.0
• Affects Version/s: Lustre 2.6.0
• Severity: 3
• Assignee: Mikhail Pershin
• Reporter: Alexey Lyashkov

Description

Lustre: DEBUG MARKER: == sanity test 36f: utime on file racing with OST BRW write ============ 12:53:21 (1389689601)
LustreError: 10989:0:(fail.c:133:__cfs_fail_timeout_set()) cfs_fail_timeout id 214 sleeping for -1000ms
schedule_timeout: wrong timeout value fffffffffffffc18
Pid: 10989, comm: ll_ost_io01_000 Tainted: G --------------- T 2.6.32-358.18.1-lustre #0
Call Trace:
[<ffffffff81532689>] ? schedule_timeout+0x2f9/0x310
[<ffffffffa03fa9c1>] ? libcfs_debug_msg+0x41/0x50 [libcfs]
[<ffffffffa03f5d9f>] ? __cfs_fail_timeout_set+0xcf/0x150 [libcfs]
[<ffffffffa0815fa9>] ? cfs_fail_timeout_set.clone.2+0x29/0x30 [ptlrpc]
[<ffffffffa081c56b>] ? tgt_brw_write+0x34b/0x1540 [ptlrpc]
[<ffffffffa0552bb3>] ? class_connected_export+0x33/0x80 [obdclass]
[<ffffffffa07b9aac>] ? lustre_msg_get_opc+0x9c/0x110 [ptlrpc]
[<ffffffffa081b995>] ? tgt_request_handle+0x245/0xad0 [ptlrpc]
[<ffffffffa07cc261>] ? ptlrpc_main+0xce1/0x1960 [ptlrpc]
[<ffffffffa07cb580>] ? ptlrpc_main+0x0/0x1960 [ptlrpc]
[<ffffffff810996b6>] ? kthread+0x96/0xa0
[<ffffffff8100c1ca>] ? child_rip+0xa/0x20
[<ffffffff81534750>] ? _spin_unlock_irq+0x30/0x40
[<ffffffff8100bb10>] ? restore_args+0x0/0x30
[<ffffffff81099620>] ? kthread+0x0/0xa0
[<ffffffff8100c1c0>] ? child_rip+0x0/0x20
LustreError: 10989:0:(fail.c:137:__cfs_fail_timeout_set()) cfs_fail_timeout id 214 awake
Lustre: DEBUG MARKER: cancel_lru_locks osc start

Issue Links

• duplicates: LU-4457 Interop 2.5.0<->2.6 failure on test suite sanity test_118d: Multiop failed to block on fsync
• is duplicated by: LU-1921 fail_loc should be restored when subtest fails in replay-ost-single
• is related to: LU-793 Reconnections should not be refused when there is a request in progress from this client.