-
Notifications
You must be signed in to change notification settings - Fork 8.2k
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.
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
[Es Archiver][Load Action] High Water Mark to 5000, Concurrency 4; Override-able #167993
Conversation
@elasticmachine merge upstream |
1 similar comment
@elasticmachine merge upstream |
51c2cd0
to
6bfb9b1
Compare
fe67122
to
2b885fb
Compare
b9f0d5d
to
336a21d
Compare
8940657
to
050e4ab
Compare
@elasticmachine merge upstream |
…ode-chages-without-benchmark-tool
/ci |
…without-benchmark-tool' of github.com:wayneseymour/kibana into esarch-fixed-higher-watermark-impl-minimal-code-chages-without-benchmark-tool
/ci |
/ci |
/ci |
/ci |
@elasticmachine merge upstream |
…ode-chages-without-benchmark-tool
/ci |
…without-benchmark-tool' of github.com:wayneseymour/kibana into esarch-fixed-higher-watermark-impl-minimal-code-chages-without-benchmark-tool
@elasticmachine merge upstream |
…ode-chages-without-benchmark-tool
/ci |
💔 Build FailedFailed CI Steps
Test Failures
Metrics [docs]Public APIs missing comments
History
To update your PR or re-run it, just comment with: |
Closing in order to have less reviewers, in a new PR. |
NEW CHANGE COMING
Summary
To speed up the load action in general, and in serverless in particular, this pr:
Increases stream batch size from
300
to5000
.Also, it's now using concurrency from the bulk call api, set to
4
.Summary of Performance Characteristics
Using a list of over 70 archives from within the repo, we've benchmarked the speed before changing the performance characteristics, and after.
One champion example is the
x-pack/test/functional/es_archives/getting_started/shakespeare
archive.It's load time decreased from ~2.5 minutes, to less than 18 seconds, in a serverless environment.
There are a little over a handful where the time actually got worse, but most are only milliseconds slower.
None are a full second slower.