ZOOKEEPER-4741: Modernize PrometheusMetricsProvider with Jetty and ne… #2291
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
dukelion
force-pushed
the
ModernPrometheusClient
branch
2 times, most recently
from
4f1edf8 to
c4c1c26
Compare
dukelion
force-pushed
the
ModernPrometheusClient
branch
3 times, most recently
from
0ae14e7 to
5b39d5c
Compare
|
@kezhuw @anmolnar @eolivelli please have a look. I'm not sure why CI is failing. UPD: just rebasing it fixed the test |
…w client
In production environments under heavy load, the existing
PrometheusMetricsProvider can introduce high latency. This is largely
due to its use of the outdated `io.prometheus.client` (simpleclient 0.x)
library and its basic, built-in HTTPServer, which has limitations in
server configuration and threading.
This commit modernizes the entire component to address these
performance and maintainability issues by:
1. **Upgrading to the `io.prometheus.metrics` (client_java 1.x) library.**
This aligns the provider with the current standard for Prometheus
instrumentation in Java and ensures future compatibility.
2. **Replacing the legacy server and threading model with an embedded
Jetty server.** The previous implementation used a custom thread
pool for a metrics processing task queue, which is now obsolete in
the new client library. This has been replaced with a robust Jetty
server, which uses its own configurable thread pool to handle
exporter servlet requests directly. This change improves stability
and simplifies the threading model, resolving the latency issues
under load.
These changes make the PrometheusMetricsProvider more stable,
performant, maintainable, and easier to configure for production use.
This resolves CI pipeline failures due to java.io.IOException: DerInputStream.getLength(): lengthTag=109, too big
dukelion
force-pushed
the
ModernPrometheusClient
branch
from
5b39d5c to
0d2ab08
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
…w client
In production environments under heavy load, the existing PrometheusMetricsProvider can introduce high latency. This is largely due to its use of the outdated
io.prometheus.client(simpleclient 0.x) library and its basic, built-in HTTPServer, which has limitations in server configuration and threading.This commit modernizes the entire component to address these performance and maintainability issues by:
Upgrading to the
io.prometheus.metrics(client_java 1.x) library. This aligns the provider with the current standard for Prometheus instrumentation in Java and ensures future compatibility.Replacing the legacy server and threading model with an embedded Jetty server. The previous implementation used a custom thread pool for a metrics processing task queue, which is now obsolete in the new client library. This has been replaced with a robust Jetty server, which uses its own configurable thread pool to handle exporter servlet requests directly. This change improves stability and simplifies the threading model, resolving the latency issues under load.
These changes make the PrometheusMetricsProvider more stable, performant, maintainable, and easier to configure for production use.