[LU-7434] lost bulk leads to a hang

• Type: Bug
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.9.0
• Labels: llnlfixready patch
• Severity: 3
• Assignee: Jinshan Xiong (Inactive)
• Reporter: Vitaly Fertman

Description

The reverse order of request_out_callback() and reply_in_callback() puts the RPC into UNREGISTERING state, which is waiting for RPC & bulk md unlink, whereas only RPC md unlink has been called so far. If bulk is lost, even expired_set does not check for UNREGISTERING state.

The same for write if server returns an error.

Issue Links

• is duplicated by: LU-9861 Client not reconnecting to OST
• is related to: LU-2522 conf-sanity test_23b timed out during mount: RPCs in "Unregistering" phase found
• is related to: LU-8062 recovery-small test_115b: @@@@@@ FAIL: dd success
• is related to: LU-8511 mdc stuck in EVICTED state
• is related to: LU-8067 sanityn test_31a: read dd: reading `/mnt/lustre2/d31a.sanityn/f31a.sanityn': Input/output error 0 blocks, must be