Summary

Telegram provider silently stops receiving messages after long-polling times out. Requires gateway restart to recover.

Symptoms

• Telegram messages stop arriving
• No errors visible in main log
• pending_update_count accumulates in Telegram API (verified via getWebhookInfo)

Cause

getUpdates times out after 500 seconds and the provider exits without auto-reconnecting.

Evidence

From ~/.openclaw/logs/gateway.err.log:

2026-02-02T00:23:50.753Z [telegram] [default] channel exited: Request to 'getUpdates' timed out after 500 seconds
2026-02-02T04:22:05.051Z [telegram] [default] channel exited: Request to 'getUpdates' timed out after 500 seconds

Workaround

Restart gateway:

launchctl stop ai.openclaw.gateway && sleep 2 && launchctl start ai.openclaw.gateway

Suggested Fix

Auto-reconnect on timeout with exponential backoff, similar to how websocket reconnection typically works.

Environment

• OpenClaw version: 2026.1.29
• macOS, launchctl-managed gateway
• Telegram channel with long-polling (no webhook)