-
Notifications
You must be signed in to change notification settings - Fork 44
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
Export tokio runtime metrics via prometheus #1524
Conversation
Test Results 99 files ±0 99 suites ±0 8m 26s ⏱️ -13s Results for commit 84cc3cf. ± Comparison against base commit bb0645f. This pull request removes 1 and adds 1 tests. Note that renamed tests count towards both.
♻️ This comment has been updated with latest results. |
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.
The more metrics the merrier. LGTM. +1 for merging.
84cc3cf
to
868573f
Compare
This PR includes: - Attachment of PPM to cluster controller now use Networking - PPM observes partition processor status through a buffered watch mechanism - PPM can now send control messages (unused at this PR) to processor for future use. - PPM collects state information from running processors. This will be used in a follow PR to response to controller requests about partitions. Cluster controller grpc service is kept for external tooling integration (CLI, etc.)
ClusterState is a new data structure maintained by ClusterController. ClusterState is established by a scatter-gather request (GetProcessorsState) to all workers. For every worker, partition processor manager is the entity responsible for responding to GetProcessorsState requests, it keeps taps on running partition processors via a set of tokio watches. This PR is a snapshot in time of progress done on this component, this is far from complete but it's best to merge the current progress until we circle back to it in a week or two.
This also changes the loglet API to accept Bytes to avoid duplication of serialization across loglet implementations. This holds the current invariant that payloads are opaque to loglets.
Export tokio runtime metrics via prometheus
Stack created with Sapling. Best reviewed with ReviewStack.