[LU-2424] add memory limits for ptlrpc service Created: 04/Dec/12  Updated: 02/Dec/13  Resolved: 15/Oct/13

Status: Resolved
Project: Lustre
Component/s: None
Affects Version/s: None
Fix Version/s: Lustre 2.4.0, Lustre 2.5.0

Type: Improvement Priority: Minor
Reporter: Andriy Skulysh Assignee: Liang Zhen (Inactive)
Resolution: Fixed Votes: 0
Labels: patch

Issue Links:
Related
is related to LU-2432 ptlrpc_alloc_rqbd spinning on vmap_ar... Resolved
is related to LU-1431 Support for larger than 1MB sequentia... Resolved
Rank (Obsolete): 5748

 Description   

Improve memmory allocation in case of huge request queues.

Limit number of request buffers in backlog, to avoid too expensive
memory using with LNet routers.
Allocate requests using cfs_mem_cache.
Allow only 1 thread to allocate request buffers.

 Comments   
Comment by Andriy Skulysh [ 04/Dec/12 ]

Xyratex-bug-id: MRP-689

Comment by Andriy Skulysh [ 04/Dec/12 ]

PATCH: http://review.whamcloud.com/#change,4740

Comment by Liang Zhen (Inactive) [ 12/Dec/12 ]

Please check discussion on LU-2432, we probably should fix memory pressure by have bigger request buffer.

Comment by Liang Zhen (Inactive) [ 01/Jan/13 ]

patch is here: http://review.whamcloud.com/#change,4940

Comment by Andriy Skulysh [ 03/Jul/13 ]

replaced previous patch with request allocation using kmem_cache only http://review.whamcloud.com/#/c/6874/ .

Comment by Jodi Levi (Inactive) [ 15/Oct/13 ]

Closing ticket as patches have landed to Master. Please let me know if more work is needed in this ticket and I will reopen.

Comment by Jian Yu [ 30/Nov/13 ]

replaced previous patch with request allocation using kmem_cache only http://review.whamcloud.com/#/c/6874/ .

The patch was back-ported to Lustre b2_4 branch: http://review.whamcloud.com/8445

Generated at Sat Feb 10 01:25:05 UTC 2024 using Jira 9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c.