[LU-1063] speed up kms recalculation

• Status: Resolved
• Project: Lustre
• Affects Version/s: Lustre 1.8.x (1.8.0 - 1.8.5)
• Type: Improvement
• Priority: Minor
• Reporter: Iurii Golovach (Inactive)
• Assignee: Cliff White (Inactive)
• Resolution: Won't Fix
• Labels: patch, performance
• Story Points: 30
• Bugzilla ID: 22,060

Description

with the patch above the kms value calculated more quickly because of next improvements:
1) interval tree was used for the going through the list of locks instead the direct linked list
search.
2) skip the subtree when maximum value in this subtree is less then new kms
3) kms search stopped when lock with size equal to the old kms value found.

this optimization shows speed increase from 20 to 40% according to the tests results.

https://bugzilla.lustre.org/show_bug.cgi?id=22060 - contains the whole story

Comments

[Iurii Golovach (Inactive)]

patch is in reviewboard
git push ssh://[e-mail]:29418/fs/lustre-release HEAD:refs/for/b1_8

[Iurii Golovach (Inactive)]

previous comment is wrong. link on the patch in the reviewboard:
http://review.whamcloud.com/#change,2076

[Peter Jones]

Iurii

Is this change also required for master? If so, we would prefer to land it there first. Could you look into porting it?

Thanks

Peter

[Andreas Dilger]

Iurii,
this feature looks very interesting and useful under some workloads. However, we are not landing features or significant code restructuring onto maintenance branches. New features like this should be targetted at a feature release like the upcoming 2.3 release.

[Iurii Golovach (Inactive)]

there is a kernel panic observed during the port into 2.2
ticket MRP-483 was created to track this issue.

[Keith Mannthey (Inactive)]

Has a Master patch been developed for this? Is this issue abandoned?

[Nathan Rutman]

Xyratex-bug-id: MRP-242

[Keith Mannthey (Inactive)]

Lurii or Nathan, is there going to be a patch for Master?

[Iurii Golovach (Inactive)]

Hi Keith,

I adopted this patch to master, but it fails on few tests. Since the bug (fixed after this optimization) was critical for 1.8 and doesn't exist in 2.x there is no rush with it.

Answer on your question - yes, there will be a patch for master, but since it's minor priority I'll create it when I have a free time for it.

[Cliff White (Inactive)]

If there is not going to be a patch for master, we should close this issue, please advise.

[Cliff White (Inactive)]

I am going to close this issue for now. If there is a new patch or any further updates please re-open.