Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Remove spinlock in MetaClient #2786

Closed
critical27 opened this issue Sep 2, 2021 · 6 comments
Closed

Remove spinlock in MetaClient #2786

critical27 opened this issue Sep 2, 2021 · 6 comments
Assignees
@CPWstatic
Labels
priority/hi-pri Priority: high type/enhancement Type: make the code neat or more efficient

Comments

@critical27
Copy link
Contributor

critical27 commented Sep 2, 2021 •
edited
Loading

Bottleneck when compaction. getTagSchema/getEdgeSchema. Use either threadlocal or rcu will be fine.
@critical27 critical27 added the type/enhancement Type: make the code neat or more efficient label Sep 2, 2021
@critical27 critical27 changed the title Remove getTagSchema/getEdgeSchema and so on spinlock in MetaClient Remove spinlock in MetaClient Sep 2, 2021
@jamieliu1023 jamieliu1023 mentioned this issue Sep 4, 2021
Closed
@cangfengzhs
Copy link
Contributor

cangfengzhs commented Sep 5, 2021 •
edited
Loading

Writing will still block reading when using threadlocal, RCU may be a better choice.And using RCU with doublebuffer may further improve performance by reducing cache line invalidation.

@critical27
Copy link
Contributor Author

Talked offline. It is ok to use a threadlocal variable in this case. We only need to block when cache need to update, which is rare case.

@CPWstatic
Copy link
Contributor

We had a threadlocal implementation in nebula1.0.

@critical27 critical27 mentioned this issue Sep 8, 2021
Merged
@critical27 critical27 added the priority/hi-pri Priority: high label Sep 8, 2021
@critical27
Copy link
Contributor Author

It seems it is a great bottleneck when rebuild index. Wait vesoft-inc/nebula-common#440 move to nebula

@Sophie-Xie
Copy link
Contributor

@critical27 Will it be resolved in v2.6.0?

@critical27
Copy link
Contributor Author

addressed in 3.0

@jamieliu1023 jamieliu1023 mentioned this issue Feb 12, 2022
Closed
yixinglu pushed a commit to yixinglu/nebula that referenced this issue Sep 14, 2023
@Aiee @Sophie-Xie
Change the LM check period back (vesoft-inc#2786)
e975ec0 
Change the check period back

Co-authored-by: Sophie <84560950+Sophie-Xie@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@CPWstatic CPWstatic

Labels
priority/hi-pri Priority: high type/enhancement Type: make the code neat or more efficient
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
4 participants
@CPWstatic @critical27 @cangfengzhs @Sophie-Xie