-
Couldn't load subscription status.
- Fork 664
Rest utils metrics #23
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
Conversation
Most importantly, this changes the default prefix for the metrics. However, the bulk of the change makes minor adjustments to work with the common-utils Time interface.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@ewencp To speed up review, could you post the jmx metric name that is created as a result of this annotation. This example should suffice to understand the rest.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Sure. This generates:
- brokers.list.request-rate
- brokers.list.request-byte-rate
- brokers.list.request-size-avg
- brokers.list.request-size-max
- brokers.list.response-rate
- brokers.list.response-byte-rate
- brokers.list.response-size-avg
- brokers.list.response-size-max
- brokers.list.request-latency-avg
- brokers.list.request-latency-max
- brokers.list.request-error-rate
That's in addition to the global ones for Jersey (same as above, but without the brokers.list prefix) and global ones for Jetty:
- connections-accepted-rate
- connections-opened-rate
- connections-closed-rate
- connections-active
All the Jersey ones are collected under a "jersey-metrics" group and Jetty ones under a "jetty-metrics" group.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@ewencp These look good. What do you think about 90, 99 and 999 percentiles for some of these, like for example the latency and byte-rate ones. One potential risk is that we don't know how well the percentiles work in common-metrics.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That's actually exactly why I didn't add them in :) The new producer code only seemed to use avg and max, so I stuck with that. But I agree, 90, 99, and 999 would be great.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ok. Let's just leave it then and come back to it when we are slightly more confident of the quality of the percentiles :)
Other than that, this patch looks good.
Conflicts: src/main/java/io/confluent/kafkarest/KafkaRestConfig.java
Add support for metrics generated automatically by rest-utils. This required a few updates due to changes to the framework. Blocked by confluentinc/rest-utils#6