-
Notifications
You must be signed in to change notification settings - Fork 2.2k
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
Add metrics to the Prometheus API #1623
Conversation
It's been two weeks now, could you please confirm this? |
@netfs Please check and merge this PR? |
please do look at c317582 that adds runtime (latency) metrics for predict/regress/classify methods (and these are exported as prometheus metrics, if you enable prometheus export). the above commit does not include latency stats for rpc/http handling, and thats where your PR would be useful. |
Conflicts: tensorflow_serving/model_servers/http_rest_api_handler.cc tensorflow_serving/servables/tensorflow/util.cc tensorflow_serving/servables/tensorflow/util.h
@netfs I fixed. |
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.
thanks for the recent changes. few more suggestions.
@netfs I fixed. |
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.
also please fix broken test:
https://source.cloud.google.com/results/invocations/efea5490-a7cc-4556-a1ac-115b7c25a0d9/targets
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.
Thanks for the change! and for being patient and addressing all my comments.
Will merge shortly. |
Sorry, I fixed test. (I don't know the command to run unit tests in my development environment.) |
Hello.
This patch adds latency and availability metrics for Predict method in gRPC API to the Prometheus API.
The metrics are measured for the entire process and for each model.
The added metric items are as follows.