Skip to content

fix: fit the RPC retry schedule inside a worker job - #695

Merged
MoonBoi9001 merged 1 commit into
mainfrom
mb9/fit-rpc-retry-schedule-inside-worker-job
Sep 17, 2026
Merged

MoonBoi9001 merged 1 commit into
mainfrom
mb9/fit-rpc-retry-schedule-inside-worker-job

Conversation

@MoonBoi9001

@MoonBoi9001 MoonBoi9001 commented Sep 17, 2026

Copy link
Copy Markdown
Member

This PR raises the worker job timeout from 300 s to 420 s and budgets a chain submission at 2 walks of the RPC provider ring (a nonce read and a send) instead of 4, so 3 providers at the default 10 s timeout and 3 retries fit inside the submission deadline. Before, 2 or more providers overran it, warned at startup, and could give up before the last provider. The health threshold moves from 600 s to 840 s.

A transaction submission may hold the sending lock for 4/5 of the 300 s job timeout, but the
retry schedule it budgeted (4 walks of the provider ring) overran that with 2 providers at the
defaults, so dipper warned at startup and could give up early. Budget 2 walks and allow 420 s.
@MoonBoi9001
MoonBoi9001 merged commit f40a71f into main Sep 17, 2026
11 checks passed
@MoonBoi9001
MoonBoi9001 deleted the mb9/fit-rpc-retry-schedule-inside-worker-job branch September 17, 2026 14:19
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant