fix: Widen wool keepalive margin to stop GOAWAY too_many_pings#73
Merged
Conversation
The worker dispatch channel ran on wool's default gRPC options, which set the client keepalive cadence (keepalive_time_ms=30s) exactly equal to the server's no-data ping floor (http2_min_recv_ping_interval_without_data_ms=30s) with a 2-strike budget. A workflow dispatch is a long-lived stream-stream RPC that goes quiet during a subprocess stage (samtools/sort on a large file); with keepalive_permit_without_calls on, the client keeps pinging into that silence. Over Fargate's awsvpc ENI, inter-ping jitter lands pings a hair under the 30s floor, the server counts strikes, and after three it sends GOAWAY too_many_pings — surfacing on the API as UNAVAILABLE and failing the job. Give the cadence real margin via a shared worker_grpc_options(): client pings once a minute, server floor drops to 20s, strike budget to 5. A ping can now arrive 3x early and still clear the floor. Wired into both worker entrypoints (worker_main for ECS, worker_lan for local) from one config point; the worker advertises the channel options to clients via discovery metadata, so both directions stay consistent.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The worker dispatch channel ran on wool's default gRPC options, which set the client keepalive cadence (keepalive_time_ms=30s) exactly equal to the server's no-data ping floor
(http2_min_recv_ping_interval_without_data_ms=30s) with a 2-strike budget. A workflow dispatch is a long-lived stream-stream RPC that goes quiet during a subprocess stage (samtools/sort on a large file); with keepalive_permit_without_calls on, the client keeps pinging into that silence. Over Fargate's awsvpc ENI, inter-ping jitter lands pings a hair under the 30s floor, the server counts strikes, and after three it sends GOAWAY too_many_pings — surfacing on the API as UNAVAILABLE and failing the job.
Give the cadence real margin via a shared worker_grpc_options(): client pings once a minute, server floor drops to 20s, strike budget to 5. A ping can now arrive 3x early and still clear the floor. Wired into both worker entrypoints (worker_main for ECS, worker_lan for local) from one config point; the worker advertises the channel options to clients via discovery metadata, so both directions stay consistent.