Recovery for locations stuck after failed refresh#33642

Merged

gibsondan merged 16 commits into

dagster-io:masterfrom

jmoldow:jordan/server-watcher-error-recovery

Jun 30, 2026

Commits on Mar 19, 2026

Recovery for locations stuck after failed refresh

jmoldow
and
claude
committed
Address PR review feedback: remove retry loop, add outer backoff, fix test timeouts

jmoldow
and
claude
committed

Commits on Jun 23, 2026

Merge branch 'master' into jordan/server-watcher-error-recovery
jmoldow
committed
feat(server_watcher): scope recovery to DagsterUserCodeUnreachableError and direct-refresh
jmoldow
committed
docs(server_watcher): clarify reconnect-loop comments and docstring
jmoldow
committed
docs(test_watch_server): clarify polling timeout and startup-reset comments
jmoldow
committed
test(server_watcher): de-flake test_grpc_watch_thread_recovery_when_errored[20]
jmoldow
committed
test(server_watcher): restore post-recovery event-count equality assertion
jmoldow
committed
refactor(server_watcher): rename set_server_id to update_server_id and tighten logging
jmoldow
committed
refactor(server_watcher): tighten review-iteration comments and add test type annotations
jmoldow
committed
test(test_watch_grpc_server): tighten recovery test and add post-on_error recovery test
jmoldow
committed
docs(server_watcher): clarify that recovery is scoped to DagsterUserCodeUnreachableError
jmoldow
committed
test(test_watch_server) - change Counter to dict
jmoldow
committed

Commits on Jun 24, 2026

Revert redundant _location_state_events_handler recovery change

Daniel Gibson
and
claude
committed
Move stuck-location recovery into a separate periodic check

Daniel Gibson
and
claude
committed

Commits on Jun 29, 2026

Merge pull request #1 from dagster-io/gibsondan/server-watcher-recovery-redesign
jmoldow
authored