fix(gateway): prevent reconnect watcher wedge after network-loss fatal error (#70344) - #70502
Closed
webtecnica wants to merge 2 commits into
Closed
fix(gateway): prevent reconnect watcher wedge after network-loss fatal error (#70344)#70502webtecnica wants to merge 2 commits into
webtecnica wants to merge 2 commits into
Conversation
added 2 commits
July 23, 2026 23:13
…asks (NousResearch#69787) Docs promise board default_workdir is inherited by every new task, but CLI and tool-created tasks defaulted to 'scratch' workspace regardless of the board setting. The dashboard was the only path that honoured default_workdir. Root cause: - hermes_cli/kanban.py --workspace default='scratch' always passed an explicit scratch kind, bypassing the board default_workdir - hermes_cli/kanban_db.py create_task only resolved default_workdir for already-persistent kinds (dir/worktree), never for the default scratch - tools/kanban_tools.py kanban_create forced workspace_kind='scratch' when unset, before the board resolution could fire Fix (three files): 1. kanban.py: --workspace default changed from 'scratch' to None; _parse_workspace_flag returns (None, None) for unset values 2. kanban_db.py: create_task resolves workspace_kind from the board's default_workdir when workspace_kind=None, using the same logic as the dashboard's _default_workspace_kind: git toplevel → worktree, plain dir → dir, no default_workdir → scratch 3. kanban_tools.py: removed the unconditional 'scratch' fallback, letting workspace_kind=None flow through to create_task for board resolution; fixed str(None) footgun in create_task call Explicit --workspace scratch preserves current behavior.
…l error (NousResearch#70344) Three-part fix for the gateway going silently deaf after a retryable fatal adapter error (e.g. httpx.ConnectError on Telegram): 1. **Detach-on-timeout in _connect_adapter_with_timeout** — Replaced plain asyncio.wait_for with the task-detach pattern used by _await_adapter_cleanup_with_timeout. asyncio.wait_for cancels the overdue task but then waits for it to exit, so a connect() that catches CancelledError can block recovery forever. The detach pattern releases the runner at the deadline via consume_detached_task_result. 2. **Ensure reconnect watcher always runs after escalation** — Added _ensure_reconnect_watcher_running(), called after queueing a retryable fatal error. If the reconnect watcher task has died (exhausted restart budget, terminal exception), it is respawned so queued platforms are never permanently stranded. 3. **Faulthandler at gateway startup** — Enabled faulthandler + SIGUSR2 dump to a rotating file under HERMES_HOME/logs/ for post-mortem diagnosis of future event-loop freezes. Tests added for _ensure_reconnect_watcher_running (alive, dead, not-started, not-running), fatal-error integration (retryable calls ensure, non-retryable does not), and _connect_adapter_with_timeout (timeout raises, success returns).
alt-glitch
added
type/bug
Contributor
Collaborator
|
Merged via #71177 — your commit was cherry-picked with authorship preserved (rebase-merge). Fixes applied on top: _ensure_reconnect_watcher_running now uses _spawn_supervised (preserving crash-restart supervision), faulthandler path uses _hermes_home (respecting profile overrides), file handle stored for clean shutdown, and _connect_adapter_with_timeout uses task.result() instead of bool(result). The detach-on-timeout concern from your PR was also merged separately via #71176 (crediting @VaitaR's #70345). Thanks for the thorough issue report and fix! |
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
4 participants
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.
Problem
The gateway goes silently deaf after a retryable fatal adapter error (e.g. httpx.ConnectError on Telegram). gateway_state shows running but platforms.telegram.state never returns to connected. No messages are processed until an external restart. The reconnect watcher never starts after the fatal error escalation.
Root causes:
Changes
1. Detach-on-timeout in _connect_adapter_with_timeout
Replaced asyncio.wait_for() with the task-detach pattern from _await_adapter_cleanup_with_timeout.
2. _ensure_reconnect_watcher_running() + call from fatal handler
Ensures the reconnect watcher is always alive after queueing a retryable fatal error.
3. Faulthandler at gateway startup
faulthandler.enable() + SIGUSR2 dump to logs for post-mortem diagnosis.
Tests Added
Testing
All existing and new tests pass (59 total).
Closes #70344