fix: drop SQLite DDL from DailyStatsActor.PreStart + surface wizard crash logs (closes #925, #936)#938
Merged
Aaronontheweb merged 1 commit intonetclaw-dev:devfrom May 8, 2026
Conversation
…gs on wizard poll-timeout (closes netclaw-dev#925, netclaw-dev#936) DailyStatsActor was running CREATE TABLE IF NOT EXISTS in PreStart on a fresh on-disk SQLite db. On Windows CI cold-start (AV scan + first-time fsync) this regularly pushed the first Ask past its 3s timeout. The schema is now owned by the migration files (existing 004_daily_stats.sql plus new 005_daily_skill_usage.sql), applied unconditionally by SchemaMigrator before the actor system starts. Test mirrors production by invoking the migrator before ActorOf. HealthCheckStepViewModel's daemon-startup polling loop also flaked on Windows: when WaitForExit(1500) missed a fast-failing fake daemon, DaemonManager.Start returned success and the loop spun for the full 30s before reporting a generic "Daemon did not become ready" message, hiding the real crash reason. Two fixes: - Fast-fail when DaemonManager.GetStatus().IsRunning flips to false during polling (real production improvement: dead daemons no longer cost 30s of wall time before reporting). - After poll exhaustion, read the structured crash-log line via DaemonManager.TryReadStartupFailureFromCrashLog and surface it on the wizard, matching the immediate-fail path's UX. Also: HealthCheckStepViewModel now takes TimeProvider (constitution compliance), and the "not ready" message is a single constant.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
DailyStatsActor.PreStartno longer runs synchronous SQLite DDL. Schema is owned by migration files: existing004_daily_stats.sql+ new005_daily_skill_usage.sql, applied unconditionally bySchemaMigratorbefore the actor system starts. On Windows CI cold-start (AV scan + first-time fsync) this DDL regularly pushed the firstAskpast its 3s timeout.HealthCheckStepViewModel.StartAndPollDaemonAsyncnow (a) fast-fails whenDaemonManager.GetStatus().IsRunningflips false during the polling loop and (b) reads any structured "Daemon startup aborted: …" line from the crash log on poll-timeout, so the wizard reports the real failure reason instead of a generic "Daemon did not become ready" message after 30s of waiting.HealthCheckStepViewModelnow takesTimeProvider(constitution compliance), and the "not ready" message is a single named constant.Why the test was flaking
Both flakes were Windows-CI-specific:
Askraced the AV-scanned cold-start fsync of the SQLite db file. Issue body has the agent analysis.WaitForExit(1500)missed a fast-failing fake daemon (cmd.exe startup overhead exceeds 1.5s on a contested Windows runner),Start()returned success and the polling loop spun for 30s before reporting a generic message that hid the crash reason.The daemon-liveness check is a real production improvement, not just a test-speed tweak — if a user's daemon dies during startup, the wizard now reports it within ~1s instead of waiting 30s.
Test plan
dotnet test src/Netclaw.Daemon.Tests/Netclaw.Daemon.Tests.csproj— 504/504 passdotnet test src/Netclaw.Cli.Tests/Netclaw.Cli.Tests.csproj— 621/621 passHealthCheckStepViewModelTests.RunWithOrchestrator_PreservesSpecificStartupFailureMessage~92ms;DailyStatsActorTests.QuerySkillUsageStats…~490ms)dotnet slopwatch analyze— 0 issuespwsh ./scripts/Add-FileHeaders.ps1 -Verify— clean