Issue Station
Sign In
Tools hereToolbox

openclaw - 💡(How to fix) Fix [Bug]: Gateway event-loop starvation during local Ollama embedded runs on Windows

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…

OpenClaw gateway repeatedly reports event-loop starvation during local Ollama embedded agent runs. While this is happening, Telegram Bot API calls from OpenClaw time out, Web UI/WebSocket may disconnect, and sessions can stall or enter compaction failure loops.

This does not appear to be general Telegram connectivity: direct curl requests to Telegram work reliably from the same machine, while OpenClaw logs show fetch timers being delayed by gateway event-loop starvation.

  • OpenClaw version: v2026.5.24-beta.1
  • OS: Windows
  • GPU: AMD Radeon RX 7800 XT, 16 GB VRAM
  • Backend: Ollama
  • Main model: ollama/qwen3.5:9b
  • Context window: 40k
  • Channels enabled: Telegram bot DM + Web UI
  • Compaction model:
    • starvation was observed before changing compaction model
    • later tested qwen2.5-coder:1.5b as compaction model

Gateway liveness warnings show the Node event loop being blocked while an embedded run is active:

Root Cause

OpenClaw gateway repeatedly reports event-loop starvation during local Ollama embedded agent runs. While this is happening, Telegram Bot API calls from OpenClaw time out, Web UI/WebSocket may disconnect, and sessions can stall or enter compaction failure loops.

This does not appear to be general Telegram connectivity: direct curl requests to Telegram work reliably from the same machine, while OpenClaw logs show fetch timers being delayed by gateway event-loop starvation.

  • OpenClaw version: v2026.5.24-beta.1
  • OS: Windows
  • GPU: AMD Radeon RX 7800 XT, 16 GB VRAM
  • Backend: Ollama
  • Main model: ollama/qwen3.5:9b
  • Context window: 40k
  • Channels enabled: Telegram bot DM + Web UI
  • Compaction model:
    • starvation was observed before changing compaction model
    • later tested qwen2.5-coder:1.5b as compaction model

Gateway liveness warnings show the Node event loop being blocked while an embedded run is active:

Code Example

[diagnostic] liveness warning: reasons=event_loop_delay,event_loop_utilization,cpu interval=50s eventLoopDelayP99Ms=21156.1 eventLoopDelayMaxMs=21156.1 eventLoopUtilization=1 cpuCoreRatio=1.001 active=1 waiting=0 queued=0 recentPhases=channels.telegram.start-account-handoff:0ms,channels.telegram.list-accounts:0ms,channels.telegram.is-configured:0ms,channels.telegram.runtime:0ms,channels.telegram.approval-bootstrap:0ms,channels.telegram.start-account-handoff:0ms work=[active=agent:main:direct:7194844943(processing/embedded_run,q=1,age=53s last=embedded_run:started)]
RAW_BUFFERClick to expand / collapse

Bug type

Behavior bug (incorrect output/state without crash)

Beta release blocker

Yes

Summary

OpenClaw gateway repeatedly reports event-loop starvation during local Ollama embedded agent runs. While this is happening, Telegram Bot API calls from OpenClaw time out, Web UI/WebSocket may disconnect, and sessions can stall or enter compaction failure loops.

This does not appear to be general Telegram connectivity: direct curl requests to Telegram work reliably from the same machine, while OpenClaw logs show fetch timers being delayed by gateway event-loop starvation.

  • OpenClaw version: v2026.5.24-beta.1
  • OS: Windows
  • GPU: AMD Radeon RX 7800 XT, 16 GB VRAM
  • Backend: Ollama
  • Main model: ollama/qwen3.5:9b
  • Context window: 40k
  • Channels enabled: Telegram bot DM + Web UI
  • Compaction model:
    • starvation was observed before changing compaction model
    • later tested qwen2.5-coder:1.5b as compaction model

Gateway liveness warnings show the Node event loop being blocked while an embedded run is active:

Steps to reproduce

  1. Start OpenClaw Gateway
  2. Telegram Message
  3. Observe gateway logs

Expected behavior

No timeouts to telegram api

Actual behavior

Telegram timing out, session stalling

OpenClaw version

v2026.5.24-beta.1

Operating system

Windows 11

Install method

npm

Model

qwen3.5:9b

Provider / routing chain

telegram bot -> openclaw -> ollama -> qwen

Additional provider/model setup details

openclaw-diagnostics-2026-05-25T00-10-16-365Z-21868.zip

Logs, screenshots, and evidence

[diagnostic] liveness warning: reasons=event_loop_delay,event_loop_utilization,cpu interval=50s eventLoopDelayP99Ms=21156.1 eventLoopDelayMaxMs=21156.1 eventLoopUtilization=1 cpuCoreRatio=1.001 active=1 waiting=0 queued=0 recentPhases=channels.telegram.start-account-handoff:0ms,channels.telegram.list-accounts:0ms,channels.telegram.is-configured:0ms,channels.telegram.runtime:0ms,channels.telegram.approval-bootstrap:0ms,channels.telegram.start-account-handoff:0ms work=[active=agent:main:direct:7194844943(processing/embedded_run,q=1,age=53s last=embedded_run:started)]

Impact and severity

No response

Additional information

No response

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

FAQ

Expected behavior

No timeouts to telegram api

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

openclaw - 💡(How to fix) Fix [Bug]: Gateway event-loop starvation during local Ollama embedded runs on Windows