-
Notifications
You must be signed in to change notification settings - Fork 553
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
rptest: Implement flink service memory autosizing #16119
Conversation
Including HTT team as a reviewers since changes will be related to the HTT code a bit and there will be new routine to identify instance specs using its type name. cc: @ivotron, @piyushredpanda |
Is this ready for review, @savex ? |
Not yet. It needs one more function in EC2 to access instance specs. |
It is ready for review. |
5de5a2c
to
99dc902
Compare
ducktape was retried in https://buildkite.com/redpanda/redpanda/builds/44098#018d31f0-bff5-4928-847b-aedb5087a3ad ducktape was retried in https://buildkite.com/redpanda/redpanda/builds/44254#018d3e3e-88eb-4690-b03c-3247efdd6a58 |
8f06229
to
a001e8b
Compare
f7e0d45
to
6d08313
Compare
Conducted quick check on EC2
|
6d08313
to
1b8eee5
Compare
Used to get meta and instance specs for AWS/EC2. Also, metadata getter is updated to use with cluster.node.
Flink is autosized to use while node's memory. It will not consume all of it, only set the maximums. Normally it would not be beyond 10G for <5 jobs.
1b8eee5
to
5d2c5fa
Compare
Flink service should detect instance specs and scale up memory for Flink's
Job Manager
andTask Manager
processes. Also, it should start additional task managers on single node for each CPUMemory sizing strategy:
10 % for system
10 % for Job Manager
80% / vcpus for each Task Manager process
Fixes: redpanda-data/devprod#1011
Backports Required
Release Notes