Attach to cluster controller through Networking #1500
Conversation
Test Results 99 files ±0 99 suites ±0 8m 5s ⏱️ -51s Results for commit 4f22a93. ± Comparison against base commit 1d7ac91. This pull request removes 1 and adds 1 tests. Note that renamed tests count towards both.♻️ This comment has been updated with latest results. |
AhmedSoliman
changed the title
|
Linter error is on generated code, this will be fixed in the next PR in this stack. |
|
Significant changes are incoming to this PR, putting this back as WIP. |
AhmedSoliman
changed the title
There was a problem hiding this comment.
Some intermediate review comments for a718fcd. Waiting for the PR to be updated before continuing with the review.
AhmedSoliman
changed the title
AhmedSoliman
force-pushed
the
pr1500
branch
6 times, most recently
from
baeeec4 to
25896ae
Compare
|
This should be ready for review. PTAL @tillrohrmann. |
AhmedSoliman
force-pushed
the
pr1500
branch
2 times, most recently
from
2a5c6d9 to
bce7fb7
Compare
| Ok(Response::new(AttachmentResponse {})) | ||
| _request: tonic::Request<ClusterStateRequest>, | ||
| ) -> Result<tonic::Response<ClusterStateResponse>, tonic::Status> { | ||
| unimplemented!() |
There was a problem hiding this comment.
Not that anyone is calling this endpoint but maybe let's return an error instead of causing a panic in case a stranger calls it?
There was a problem hiding this comment.
It shouldn't panic the runtime, but better to change this to an error code instead.
| running_partition_processors: HashMap::default(), | ||
| running_partition_processors: BTreeMap::default(), |
There was a problem hiding this comment.
Is BTreeMap faster than HashMap for small number of entries?
There was a problem hiding this comment.
Correct, it's also more cacheline friendly, and in general better when keys are small and numeric.
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.)
Attach to cluster controller through Networking
This PR includes:
Cluster controller grpc service is kept for external tooling integration (CLI, etc.)
Stack created with Sapling. Best reviewed with ReviewStack.