[LU-12294] Memory leak after recovery packet processing Created: 13/May/19  Updated: 16/May/19

Status: Open
Project: Lustre
Component/s: None
Affects Version/s: Lustre 2.12.1
Fix Version/s: None

Type: Bug Priority: Minor
Reporter: Tatsushi Takamura Assignee: Tatsushi Takamura
Resolution: Unresolved Votes: 0
Labels: None

Severity: 3
Rank (Obsolete): 9223372036854775807

 Description   

When the driver stops, resources for recovery packets will be released.
But, If recovery process is working at this time, resources are not released.

We correct to wait for recovery packets to be unlinked.

 



 Comments   
Comment by Amir Shehata (Inactive) [ 16/May/19 ]

Could you please check if the below patches fix this issue?

https://review.whamcloud.com/#/c/34445/4
https://review.whamcloud.com/#/c/34477/5
https://review.whamcloud.com/#/c/34252/7
https://review.whamcloud.com/#/c/34607/3
https://review.whamcloud.com/#/c/34770/2
https://review.whamcloud.com/#/c/34771/2
https://review.whamcloud.com/#/c/34778/2
https://review.whamcloud.com/#/c/34796/2
https://review.whamcloud.com/#/c/34798/3
https://review.whamcloud.com/#/c/34885/1

https://review.whamcloud.com/#/c/34477/5

 

Please note there has been other fixes which had landed on master previously. I'm not sure if you pulled them in, in your testing

139d69141b73d427490f39d3096b2187e979eaea LU-11472 lnet: Decrement health on timeout
5d77f0d8dc74c752032e449687090ff1360cd32e LU-11476 lnet: set the health status correctly
b04c35874dca17a191ddbac9b52b005066a10dda LU-11283 lnet: fix setting health value manually
e0132e16df15290fde67a69e64061d31fb3e0923 LU-11474 lnet: unlink md if fail to send recovery
cf3cc2c72e6e6ee4a1d1f56c1b427b2279136416 LU-11271 lnd: conditionally set health status
a5c1cd5ec240543a3b77b5da86eaceff0ace5ba0 LU-11530 lnet: properly error check sensitivity
dc1f5f08b420aba99f613a6bc6b8acb7afd0e894 LU-11468 lnet: configure recovery interval
307d877e51240e1d50c0fc42b190f1247f387f85 LU-11469 lnet: fix "debug recovery" output
2be10428ac22426c5868b699b6c0b80c040465dc LU-11514 lnet: separate ni state from recovery
b7f8d156db696fcc15fd37cfdfbee6549148fb69 LU-11468 lnet: set recovery interval from lnetctl
a3b6109705dc7b0e2bc376e3b55c8434d0fa0a39 LU-10876 lnet: peer deletion code may hide error
8c249097e62713baf51aec808489a86acf46748d LU-11734 lnet: handle multi-md usage
Generated at Sat Feb 10 02:51:19 UTC 2024 using Jira 9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c.