[LU-10034] LDISKFS Quota scalability improvement Created: 26/Sep/17  Updated: 28/Jun/21  Resolved: 28/Jun/21

Status: Closed
Project: Lustre
Component/s: None
Affects Version/s: Lustre 2.11.0
Fix Version/s: None

Type: Improvement Priority: Minor
Reporter: Wang Shilong (Inactive) Assignee: Wang Shilong (Inactive)
Resolution: Won't Fix Votes: 0
Labels: None

Issue Links:
Related
is related to LU-12353 optimizations for ldiskfs quota updates Resolved
Rank (Obsolete): 9223372036854775807

 Description   

Currently Lustre enabled user/group accounting in default, in our testing, we found
that with these extra object accounting, there is big overhead especially for metadata
intensive workload, there will be more overhead with project quota enabled.(Numbers
will be attached later)

Ideally, we could reach similar performance number with/without quota.

There are some Kernel and Ext4 patches merged by Linux upstream which could
significantly improve the quota performance, it will be good to have them in Lustre
RHEL7.



 Comments   
Comment by Peter Jones [ 26/Sep/17 ]

This sounds like something you are working on yourself

Generated at Sat Feb 10 02:31:27 UTC 2024 using Jira 9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c.