#V.2026.4.26 Response issue

1 messages · Page 1 of 1 (latest)

trail shell
#

Ever since I updated to this version, I wont get any response on neither WhatsApp or Telegram. And the main channel I use is normally Discord, and there I can see the agent typing, then it stops, then it starts again, then it takes between 3 - 10 minutes before I actually get a response. Anyone knows if this is a common issue with the new patch(es) and is there a fix?

wet ravineBOT
#

To help others find answers, you can mark your question as solved via Right click solution message -> Apps -> ✅ Mark Solution

twilit talonBOT
#

Yeah, this sounds like a gateway/agent runtime stall, not a pure WhatsApp/Telegram config issue — especially since Discord shows typing but replies arrive 3–10 min later.

First thing I’d do:

openclaw --version
openclaw update
openclaw doctor
openclaw gateway restart
openclaw channels status --probe

openclaw update is the recommended update path and runs doctor/restarts as part of the flow; 2026.4.26 is currently shown as the latest GitHub release. (docs.openclaw.ai)

If it still happens, please paste:

# 
openclaw status --all
openclaw channels status --probe

and a log snippet from right after sending one Discord/Telegram message:

openclaw logs --follow

Things I’d look for: provider timeouts/fallbacks, compaction, stuck tool-result delivery, Telegram polling stalls, WhatsApp disconnects, or session lock/queue messages. Docs for the channel triage ladder are here: https://docs.openclaw.ai/channels/troubleshooting (docs.openclaw.ai)

If this started exactly after 2026.4.26 and survives update/doctor/restart, it’s worth filing a GitHub issue with that output: https://github.com/openclaw/openclaw/issues

trail shell
#

I do use AWS Bedrock (Claude Sonnet/Opuse 4.6/4.7) could this be part of the issues? It did work perfectly fine before the updates though

twilit talonBOT
#

Yes — Bedrock could absolutely be part of it.

Since Discord shows typing, the inbound channel is working; the delay is likely happening inside the agent/model run. Bedrock failures/timeouts/replay issues can make all channels look broken because WhatsApp/Telegram/Discord are waiting on the same agent response.

For 2026.4.26 specifically, I’d update first — there were Bedrock/session fixes after that.

Please run:

openclaw --version
openclaw update
openclaw gateway restart
openclaw doctor

Then test Bedrock directly:

# 
openclaw infer model run --model <your amazon-bedrock model ref> --prompt "Reply with one short sentence." --json

And if it’s still slow, grab logs while sending one Discord message:

openclaw logs --follow

Look for Bedrock/AWS errors like validation failures, throttling, stream timeouts, credential/region issues, or fallback retries.

Bedrock docs: https://docs.openclaw.ai/providers/bedrock
Channel triage: https://docs.openclaw.ai/channels/troubleshooting

trail shell
#

Connecting to OpenClaw EC2 at 3.234.218.33...

Welcome to Ubuntu 24.04.4 LTS (GNU/Linux 6.17.0-1012-aws x86_64)

System information as of Tue Apr 28 20:20:30 UTC 2026

System load: 0.92 Temperature: -273.1 C
Usage of /: 7.1% of 95.82GB Processes: 148
Memory usage: 12% Users logged in: 1
Swap usage: 0% IPv4 address for ens5: 172.31.68.155

Expanded Security Maintenance for Applications is not enabled.

0 updates can be applied immediately.

Enable ESM Apps to receive additional future security updates.
See https://ubuntu.com/esm or run: sudo pro status

Last login: Tue Apr 28 20:10:58 2026 from 155.4.131.84
ubuntu@ip-172-31-68-155:~$ openclaw --version
openclaw update
openclaw doctor
openclaw gateway restart
openclaw channels status --probe
OpenClaw 2026.4.26 (be8c246)
Updating OpenClaw...

│
◇ ✓ Updating via package manager (38.37s)
│
◇ ✓ Running doctor checks (41.31s)

Update Result: OK
Root: /home/ubuntu/.npm-global/lib/node_modules/openclaw
Before: 2026.4.26
After: 2026.4.26

Total time: 82.92s

Updating plugins...
Config warnings:

  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
    Config overwrite: /home/ubuntu/.openclaw/openclaw.json (sha256 d63d64a2b997a4ece4238e5e752d362682b24326c6ac160c3af1abfe9dc99a17 -> 339b391256e0dcb8e20f598695df209aa39432d8fd07c471de1814820ca85125, backup=/home/ubuntu/.openclaw/openclaw.json.bak)
    Config warnings:
  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
    npm plugins: 0 updated, 0 unchanged, 2 skipped.
    Completion cache update failed: Error: spawnSync /usr/bin/node ETIMEDOUT

Restarting service...
Gateway did not become healthy after restart.
Gateway version mismatch: expected 2026.4.26, running gateway reported unavailable.
Service runtime: status=running, state=active, pid=267018, lastExit=0
Gateway port 18789 status: free.
Restart log: /home/ubuntu/.openclaw/logs/gateway-restart.log
Run openclaw gateway status --deep for details.

Config warnings:

  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix

🦞 OpenClaw 2026.4.26 (be8c246) — I run on caffeine, JSON5, and the audacity of "it worked on my machine."

#

Config warnings:

  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix

🦞 OpenClaw 2026.4.26 (be8c246) — I don't sleep, I just enter low-power mode and dream of clean diffs.

│
◇ Config warnings ───────────────────────────────────────────────────────────────────────╮
│ │
│ - plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current │
│ bundled plugin directory; remove this redundant path or run openclaw doctor --fix │
│ - plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current │
│ bundled plugin directory; remove this redundant path or run openclaw doctor --fix │
│ │
├─────────────────────────────────────────────────────────────────────────────────────────╯
Restarted systemd service: openclaw-gateway.service
Config warnings:

  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix

🦞 OpenClaw 2026.4.26 (be8c246) — I read logs so you can keep pretending you don't have to.

│
◇ Config warnings ───────────────────────────────────────────────────────────────────────╮
│ │
│ - plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current │
│ bundled plugin directory; remove this redundant path or run openclaw doctor --fix │
│ - plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current │
│ bundled plugin directory; remove this redundant path or run openclaw doctor --fix │
│ │
├─────────────────────────────────────────────────────────────────────────────────────────╯
│
Config warnings:

  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
    ◇
    Gateway not reachable: gateway timeout after 30000ms
    Gateway target: ws://127.0.0.1:18789
    Source: local loopback
    Config: /home/ubuntu/.openclaw/openclaw.json
    Bind: loopback
    Gateway not reachable; showing config-only status.
    Config: /home/ubuntu/.openclaw/openclaw.json
    Mode: local

Tip: status --deep adds gateway health probes to status output (requires a reachable gateway).
ubuntu@ip-172-31-68-155:~$

twilit talonBOT
#

That output changes the diagnosis: your Gateway is not actually coming up cleanly.

Key line:

Service runtime: status=running ... 
Gateway port 18789 status: free.

So systemd has a Node process, but OpenClaw isn’t listening on the Gateway port. That would explain WhatsApp/Telegram not replying, and Discord being weird/slow. I wouldn’t chase Bedrock first until the Gateway startup issue is fixed.

Please run:

openclaw gateway status --deep
cat ~/.openclaw/logs/gateway-restart.log
# 
openclaw logs --tail 200

If logs point at those plugin path warnings, try:

openclaw doctor --fix
openclaw gateway restart
openclaw gateway status --deep

The Completion cache update failed: spawnSync /usr/bin/node ETIMEDOUT is suspicious too — not necessarily fatal, but it suggests startup/update work is timing out somewhere.

Also: openclaw update stayed on 2026.4.26, so you didn’t actually get a newer patch from npm yet. If gateway-restart.log shows a startup hang/crash, that’s the useful bit to paste here.

trail shell
#

ubuntu@ip-172-31-68-155:~$ openclaw gateway status --deep
cat ~/.openclaw/logs/gateway-restart.log
Config warnings:

  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
  • plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix

🦞 OpenClaw 2026.4.26 (be8c246) — I speak fluent bash, mild sarcasm, and aggressive tab-completion energy.

│
◇
Service: systemd (enabled)
File logs: /tmp/openclaw/openclaw-2026-04-28.log
Command: /usr/bin/node /home/ubuntu/.npm-global/lib/node_modules/openclaw/dist/index.js gateway --port 18789
Service file: ~/.config/systemd/user/openclaw-gateway.service
Service env: OPENCLAW_GATEWAY_PORT=18789

Config (cli): ~/.openclaw/openclaw.json
Config (service): ~/.openclaw/openclaw.json

Gateway: bind=loopback (127.0.0.1), port=18789 (service args)
Probe target: ws://127.0.0.1:18789
Dashboard: http://127.0.0.1:18789/
Probe note: Loopback-only gateway; only local clients can connect.

Runtime: running (pid 267440, state active, sub running, last exit 0, reason 0)
Warm-up: launch agents can take a few seconds. Try again shortly.
Connectivity probe: failed
Probe target: ws://127.0.0.1:18789
timeout
Capability: unknown

Port 18789 is already in use.

  • pid 267440 ubuntu: openclaw-gateway (127.0.0.1:18789)
  • Gateway already running locally. Stop it (openclaw gateway stop) or use a different port.
    Listening: 127.0.0.1:18789
    Other gateway-like services detected (best effort):
  • transcript-watcher.service (user, unit: /home/ubuntu/.config/systemd/user/transcript-watcher.service)
    Cleanup hint: systemctl --user disable --now openclaw-gateway.service
    Cleanup hint: rm ~/.config/systemd/user/openclaw-gateway.service

Recommendation: run a single gateway per machine for most setups. One gateway supports multiple agents (see docs: /gateway#multiple-gateways-same-host).
If you need multiple gateways (e.g., a rescue bot on the same host), isolate ports + config/state (see docs: /gateway#multiple-gateways-same-host).

Troubles: run openclaw status
Troubleshooting: https://docs.openclaw.ai/troubleshooting
[2026-04-28T20:23:54Z] openclaw restart attempt source=update target=openclaw-gateway.service
[2026-04-28T20:23:54Z] openclaw restart done source=update
ubuntu@ip-172-31-68-155:~$

twilit talonBOT
#

Yep — this confirms it’s not Bedrock yet. The Gateway itself is timing out before it becomes reachable:

Gateway not reachable: gateway timeout after 30000ms
Gateway target: ws://127.0.0.1:18789

The plugins.load.paths warning is probably cleanup noise; the important bit is the Gateway never starts listening.

Please paste these next:

tail -n 200 ~/.openclaw/logs/gateway-restart.log
openclaw logs --tail 200
journalctl --user -u openclaw-gateway.service -n 200 --no-pager
#

If the journalctl --user one says no unit found, try:

sudo journalctl -u openclaw-gateway.service -n 200 --no-pager

After we see the startup error/hang, we can pick the right fix. Docs for this path: https://docs.openclaw.ai/gateway/troubleshooting

#

Okay, now it’s listening on the port but not answering the Gateway probe:

Listening: 127.0.0.1:18789
Connectivity probe: failed ... timeout

That usually means the Gateway process is alive, but stuck during startup / event loop work / channel/plugin init.

Next useful thing is the actual runtime log:

tail -n 250 /tmp/openclaw/openclaw-2026-04-28.log
journalctl --user -u openclaw-gateway.service -n 200 --no-pager

Also run the cleanup the warning suggests, just to remove that variable:

# 
openclaw doctor --fix
openclaw gateway restart
sleep 60
openclaw gateway status --deep

The transcript-watcher.service is probably unrelated unless it’s hammering the Gateway, but don’t chase that yet. The log tail should show whether it’s stuck on Bedrock discovery, plugin runtime deps, channel startup, memory/QMD, etc.

trail shell
#

ubuntu@ip-172-31-68-155:~$ tail -n 200 ~/.openclaw/logs/gateway-restart.log
openclaw logs --tail 200
journalctl --user -u openclaw-gateway.service -n 200 --no-pager
[2026-04-28T20:23:54Z] openclaw restart attempt source=update target=openclaw-gateway.service
[2026-04-28T20:23:54Z] openclaw restart done source=update
error: unknown option '--tail'
Apr 28 20:20:51 ip-172-31-68-155 node[264139]: 2026-04-28T20:20:51.981+00:00 [discord] gateway: Gateway websocket closed: 1000
Apr 28 20:20:52 ip-172-31-68-155 node[264139]: 2026-04-28T20:20:52.010+00:00 [discord] gateway: Gateway reconnect scheduled in 918ms (close, resume=true)
Apr 28 20:20:58 ip-172-31-68-155 node[264139]: 2026-04-28T20:20:58.554+00:00 [ws] ⇄ res ✓ channels.status 2212ms conn=66d9e875…022b id=f67871df…d255
Apr 28 20:22:01 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:01.854+00:00 [ws] ⇄ res ✓ channels.status 558ms conn=939ad18e…0694 id=208b048d…ab4b
Apr 28 20:22:03 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:03.006+00:00 [plugins] memory-lancedb: plugin registered (db: /home/ubuntu/.openclaw/memory/lancedb, lazy init)
Apr 28 20:22:03 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:03.011+00:00 [ws] ⇄ res ✓ doctor.memory.status 1139ms conn=fa6af1a8…69a4 id=b09aaf62…a089
Apr 28 20:22:18 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:18.888+00:00 [reload] config change detected; evaluating reload (plugins.load, meta.lastTouchedVersion, meta.lastTouchedAt)
Apr 28 20:22:18 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:18.900+00:00 [reload] config change requires gateway restart (plugins.load)
Apr 28 20:22:18 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:18.905+00:00 [gateway] signal SIGUSR1 received
Apr 28 20:22:18 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:18.909+00:00 [gateway] received SIGUSR1; restarting
Apr 28 20:22:19 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:19.798+00:00 Config warnings:
Apr 28 20:22:19 ip-172-31-68-155 node[264139]: - plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
Apr 28 20:22:19 ip-172-31-68-155 node[264139]: - plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix
Apr 28 20:22:20 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:20.330+00:00 [plugins] memory-lancedb: stopped
Apr 28 20:22:20 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:20.337+00:00 [gmail-watcher] gmail watcher stopped
Apr 28 20:22:20 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:20.533+00:00 [ws] webchat disconnected code=1012 reason=service restart conn=a23aa50b-6233-48a4-b870-174ddd77e7f0
Apr 28 20:22:20 ip-172-31-68-155 node[264139]: 2026-04-28T20:22:20.542+00:00 [gateway] restart mode: full process restart (supervisor restart)
Apr 28 20:22:20 ip-172-31-68-155 systemd[1375]: openclaw-gateway.service: Consumed 9min 17.658s CPU time, 1.1G memory peak, 0B memory swap peak.

And like 24k more lines

#

ubuntu@ip-172-31-68-155:~$ sudo journalctl -u openclaw-gateway.service -n 200 --no-pager
-- No entries --
ubuntu@ip-172-31-68-155:~$

twilit talonBOT
#

Got it. Two things:

  1. openclaw logs --tail isn’t supported in your 2026.4.26 build — use openclaw logs --follow or tail the file directly.
  2. The useful part is after the restart, not the older pre-restart noise.

Run this and paste the last chunk:

journalctl --user -u openclaw-gateway.service --since "2026-04-28 20:22:20" --no-pager | tail -n 300
tail -n 300 /tmp/openclaw/openclaw-2026-04-28.log

Also do the config cleanup because plugins.load triggered a restart:

openclaw doctor --fix
# 
openclaw gateway restart
sleep 60
openclaw gateway status --deep

If it’s still timing out, try temporarily disabling startup update checks while we debug:

openclaw config set update.checkOnStart false
openclaw gateway restart

That won’t fix a model issue, but it removes one startup variable. Right now the main problem is still: process alive, port open, Gateway probe times out. That means startup/event loop is wedged somewhere, and the next log chunk should show where.

#

That’s expected — it’s a user systemd service, not a system service, so sudo journalctl -u ... won’t see it.

Use:

journalctl --user -u openclaw-gateway.service --since "2026-04-28 20:22:20" --no-pager | tail -n 300

And the direct file log:

tail -n 300 /tmp/openclaw/openclaw-2026-04-28.log

Those are the ones that should show what’s wedging the Gateway.

trail shell
#

runtimeVersion":"22.22.2","hostname":"unknown","name":"openclaw","date":"2026-04-28T20:29:28.785Z","logLevelId":3,"logLevelName":"INFO","path":{"fullFilePath":"file:///home/ubuntu/.npm-global/lib/node_modules/openclaw/dist/console-D6VW2Y3e.js:147:46","fileName":"console-D6VW2Y3e.js","fileNameWithLine":"console-D6VW2Y3e.js:147","fileColumn":"46","fileLine":"147","filePath":".npm-global/lib/node_modules/openclaw/dist/console-D6VW2Y3e.js","filePathWithLine":".npm-global/lib/node_modules/openclaw/dist/console-D6VW2Y3e.js:147","method":"console.log"}},"time":"2026-04-28T20:29:28.785+00:00","hostname":"ip-172-31-68-155","message":"▄

twilit talonBOT
#

That snippet is just a cut-off JSON log record for a console/progress character:

"message":"▄

So not the actual error yet — the useful lines are probably right before/after it.

Run this to extract the interesting bits from the giant log:

grep -Ei "error|warn|fail|timeout|stuck|fatal|exception|uncaught|startup|ready|listening|bedrock|discord|telegram|whatsapp|plugin|memory|lancedb|qmd" /tmp/openclaw/openclaw-2026-04-28.log | tail -n 200

And this for the service log, without sudo:

# 
journalctl --user -u openclaw-gateway.service --since "2026-04-28 20:22:20" --no-pager -o cat | tail -n 300

If that still only shows spinner/progress junk, do a foreground startup trace:

openclaw gateway stop
OPENCLAW_GATEWAY_STARTUP_TRACE=1 openclaw gateway --port 18789 --verbose

Let it sit until it either says ready or hangs for ~60–90s, then paste the last visible section. That should show exactly which startup phase is wedging.

trail shell
#

:28:53.437+00:00","hostname":"ip-172-31-68-155","message":"bonjour: advertised gateway fqdn=ip-172-31-68-155 (OpenClaw)._openclaw-gw._tcp.local. host=ip-172-31-68-155.local. port=18789 state=unannounced"}
{"0":"{"subsystem":"plugins"}","1":"memory-lancedb: plugin registered (db: /home/ubuntu/.openclaw/memory/lancedb, lazy init)","_meta":{"runtime":"node","runtimeVersion":"22.22.2","hostname":"unknown","name":"{"subsystem":"plugins"}","parentNames":["openclaw"],"date":"2026-04-28T20:29:13.529Z","logLevelId":3,"logLevelName":"INFO","path":{"fullFilePath":"file:///home/ubuntu/.npm-global/lib/node_modules/openclaw/dist/subsystem-rHhUC6qs.js:147:14","fileName":"subsystem-rHhUC6qs.js","fileNameWithLine":"subsystem-rHhUC6qs.js:147","fileColumn":"14","fileLine":"147","filePath":".npm-global/lib/node_modules/openclaw/dist/subsystem-rHhUC6qs.js","filePathWithLine":".npm-global/lib/node_modules/openclaw/dist/subsystem-rHhUC6qs.js:147","method":"logToFile"}},"time":"2026-04-28T20:29:13.530+00:00","hostname":"ip-172-31-68-155","message":"memory-lancedb: plugin registered (db: /home/ubuntu/.openclaw/memory/lancedb, lazy init)"}
{"0":"Config warnings:\n- plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix\n- plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix","_meta":{"runtime":"node","runtimeVersion":"22.22.2","hostname":"unknown","name":"openclaw","date":"2026-04-28T20:29:28.727Z","logLevelId":4,"logLevelName":"WARN","path":{"fullFilePath":"file:///home/ubuntu/.npm-global/lib/node_modules/openclaw/dist/console-D6VW2Y3e.js:148:46","fileName":"console-D6VW2Y3e.js","fileNameWithLine":"console-D6VW2Y3e.js:148","fileColumn":"46","fileLine":"148","filePath":".npm-global/lib/node_modules/openclaw/dist/console-D6VW2Y3e.js","filePathWithLine":".npm-global/lib/node_modules/openclaw/dist/console-D6VW2Y3e.js:148","method":"console.warn"}},"time":"2026-04-28T20:29:28.732+00:00","hostname":"ip-172-31-68-155","message":"Config warnings:\n- plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix\n- plugins: plugin: ignored plugins.load.paths entry that points at OpenClaw's current bundled plugin directory; remove this redundant path or run openclaw doctor --fix"}
{"0":"{"subsystem":"plugins"}","1":"memory-lancedb: plugin registered (db: /home/ubuntu/.openclaw/memory/lancedb, lazy init)","_meta":{"runtime":"node","runtimeVersion":"22.22.2","hostname":"unknown","name":"{"subsystem":"plugins"}","parentNames":["openclaw"],"date":"2026-04-28T20:29:53.745Z","logLevelId":3,"logLevelName":"INFO","path":{"fullFilePath":"file:///home/ubuntu/.npm-global/lib/node_modules/openclaw/dist/subsystem-rHhUC6qs.js:147:14","fileName":"subsystem-rHhUC6qs.js","fileNameWithLine":"subsystem-rHhUC6qs.js:147","fileColumn":"14","fileLine":"147","filePath":".npm-global/lib/node_modules/openclaw/dist/subsystem-rHhUC6qs.js","filePathWithLine":".npm-global/lib/node_modules/openclaw/dist/subsystem-rHhUC6qs.js:147","method":"logToFile"}},"time":"2026-04-28T20:29:53.745+00:00","hostname":"ip-172-31-68-155","message":"memory-lancedb: plugin registered (db: /home/ubuntu/.openclaw/memory/lancedb, lazy init)"}
{"0":"\u001b[38;5;167m\u001b[1mHealth check failed\u001b[22m\u001b[39m: gateway timeout after 10000ms\n \u001b[38;5;138mGateway target:\u001b[39m \u001b[38;5;209mws://127.0.0.1:18789\u001b[39m\n \u001b[38;5;138mSource:\u001b[39m \u001b[38;5;138mlocal loopback\u001b[39m\n \u001b[38;5;138mConfig:\u001b[39m \u001b[38;5;209m/home/ubuntu/.openclaw/openclaw.json\u001b[39m\n \u001b[38;5;138mBind:\u001b[39m \u001b[38;5;216mloopback\u001b[39m","_meta":

#

2026-04-28T20:30:48.945+00:00 (node:268615) [DEP0040] DeprecationWarning: The punycode module is deprecated. Please use a userland alternative instead.
(Use node --trace-deprecation ... to show where the warning was created)
2026-04-28T20:30:48.966+00:00 [gateway] starting HTTP server...
2026-04-28T20:30:48.972+00:00 [gateway] signal SIGTERM received
2026-04-28T20:30:48.976+00:00 [gateway] received SIGTERM; restarting
2026-04-28T20:30:48.984+00:00 [canvas] host mounted at http://127.0.0.1:18789/__openclaw__/canvas/ (root /home/ubuntu/.openclaw/canvas)
2026-04-28T20:30:49.021+00:00 [gateway] restart mode: full process restart (supervisor restart)
Stopped openclaw-gateway.service - OpenClaw Gateway (v2026.4.26).
openclaw-gateway.service: Consumed 23.451s CPU time, 292.1M memory peak, 0B memory swap peak.
Started openclaw-gateway.service - OpenClaw Gateway (v2026.4.26).
2026-04-28T20:30:56.065+00:00 [gateway] loading configuration…
2026-04-28T20:30:56.280+00:00 [gateway] resolving authentication…
2026-04-28T20:30:56.285+00:00 [gateway] starting...
2026-04-28T20:31:04.789+00:00 [plugins] memory-lancedb: plugin registered (db: /home/ubuntu/.openclaw/memory/lancedb, lazy init)
2026-04-28T20:31:07.407+00:00 (node:268711) [DEP0040] DeprecationWarning: The punycode module is deprecated. Please use a userland alternative instead.
(Use node --trace-deprecation ... to show where the warning was created)
2026-04-28T20:31:07.426+00:00 [gateway] starting HTTP server...
2026-04-28T20:31:07.439+00:00 [canvas] host mounted at http://127.0.0.1:18789/__openclaw__/canvas/ (root /home/ubuntu/.openclaw/canvas)
2026-04-28T20:31:07.895+00:00 [health-monitor] started (interval: 300s, startup-grace: 60s, channel-connect-grace: 120s)
2026-04-28T20:31:08.071+00:00 [gateway] agent model: amazon-bedrock/us.anthropic.claude-sonnet-4-6
2026-04-28T20:31:08.074+00:00 [gateway] http server listening (7 plugins: active-memory, bonjour, diagnostics-prometheus, discord, memory-lancedb, telegram, tokenjuice; 11.8s)
2026-04-28T20:31:08.077+00:00 [gateway] log file: /tmp/openclaw/openclaw-2026-04-28.log
2026-04-28T20:31:08.130+00:00 [model-pricing] OpenRouter pricing fetch failed: TypeError: fetch failed
2026-04-28T20:31:08.133+00:00 [model-pricing] LiteLLM pricing fetch failed: TypeError: fetch failed
2026-04-28T20:31:13.335+00:00 [gateway] starting channels and sidecars...
2026-04-28T20:31:13.352+00:00 [plugins] bonjour: watchdog detected non-announced service; attempting re-advertise (gateway fqdn=ip-172-31-68-155 (OpenClaw)._openclaw-gw._tcp.local. host=ip-172-31-68-155.local. port=18789 state=probing)
ubuntu : PWD=/home/ubuntu ; USER=root ; COMMAND=/usr/bin/tailscale serve --bg --yes 18789
pam_unix(sudo:session): session opened for user root(uid=0) by ubuntu(uid=1000)
pam_unix(sudo:session): session closed for user root
2026-04-28T20:31:14.487+00:00 [hooks] loaded 2 internal hook handlers
2026-04-28T20:31:14.774+00:00 [tailscale] serve enabled: https://openclaw-us.tail9cf1d4.ts.net/ (WS via wss://openclaw-us.tail9cf1d4.ts.net)
ubuntu@ip-172-31-68-155:~$

twilit talonBOT
#

That log still doesn’t show the real wedge — it only shows:

  • Gateway gets far enough to advertise Bonjour.
  • memory-lancedb registers.
  • config warning repeats.
  • health check times out.

The config warning matters because earlier your log said:

config change requires gateway restart (plugins.load)

So first clean that up:

# 
openclaw doctor --fix
openclaw gateway restart
sleep 90
openclaw gateway status --deep

If it still times out, run it in the foreground with startup tracing:

openclaw gateway stop
OPENCLAW_NO_AUTO_UPDATE=1 OPENCLAW_GATEWAY_STARTUP_TRACE=1 openclaw gateway --port 18789 --verbose

Let that sit for 60–90 seconds, then paste the last section it prints. That should show which startup phase is blocking. Right now we don’t have enough signal to blame Bedrock yet — the Gateway RPC itself is hanging.

#

Good, this is progress — the Gateway is now actually starting:

[gateway] http server listening ... 11.8s
[gateway] agent model: amazon-bedrock/us.anthropic.claude-sonnet-4-6

Two important clues:

  1. WhatsApp is not loaded in that plugin list:

    active-memory, bonjour, diagnostics-prometheus, discord, memory-lancedb, telegram, tokenjuice

    So WhatsApp won’t respond until we figure out why that channel/plugin isn’t enabled or loaded.

  2. The Gateway is using Bedrock Sonnet 4.6, so if Discord/Telegram are slow now, next check is model/runtime latency.

#

Run these now:

openclaw gateway status --deep
openclaw channels status --probe
openclaw infer model run --model amazon-bedrock/us.anthropic.claude-sonnet-4-6 --prompt "Reply with OK only." --json

If the infer command takes minutes or fails, it’s Bedrock/model/auth/network.
If infer is fast but Discord/Telegram are slow, it’s channel/session queue/runtime.

Also: the OpenRouter/LiteLLM pricing fetch failures are probably harmless unless you use those providers, but they may hint at general outbound network/proxy weirdness.

trail shell
#

Its not the models themself. Im telling you the issues started once I updated to the latest version and worked fine when I downgraded two versions

twilit talonBOT
#

You’re right — if downgrading two versions makes it work again, then treat this as a 2026.4.26 regression until proven otherwise. I shouldn’t keep pushing Bedrock as the main theory.

Best move now:

# pin back to the known-good version
npm i -g openclaw@<working-version>

# prevent startup update noise while debugging
openclaw config set update.checkOnStart false
openclaw config set update.auto.enabled false

openclaw doctor
openclaw gateway restart
openclaw gateway status --deep
#

Then file this as a GitHub issue with:

  • working version
  • broken version: 2026.4.26 (be8c246)
  • EC2 / Ubuntu 24.04 / Node 22.22.2
  • provider: amazon-bedrock/us.anthropic.claude-sonnet-4-6
  • symptoms: Discord typing loop + delayed response, Telegram/WhatsApp no replies
  • log snippet showing Gateway startup and loaded plugins

Issue link: https://github.com/openclaw/openclaw/issues

The strongest clue is exactly what you said: same config/provider, old version works, new version doesn’t. That’s a regression report, not normal troubleshooting anymore.

trail shell
#

2026.4.23 was the working version

twilit talonBOT
#

Perfect, then I’d roll back to 2026.4.23 and pin it while this gets reported.

npm i -g openclaw@2026.4.23
openclaw config set update.checkOnStart false
openclaw config set update.auto.enabled false
openclaw doctor
openclaw gateway restart
openclaw gateway status --deep
openclaw channels status --probe

If npm/global path weirdness shows up, use the installer pin instead:

curl -fsSL https://openclaw.ai/install.sh | bash -s -- --install-method npm --version 2026.4.23 --no-onboard
#

Then please file this as a regression: https://github.com/openclaw/openclaw/issues

Suggested title:

Regression in 2026.4.26: Discord delayed responses, Telegram/WhatsApp no replies on EC2 with Bedrock

Include: working 2026.4.23, broken 2026.4.26 (be8c246), Ubuntu 24.04 EC2, Node 22.22.2, Bedrock Sonnet 4.6, and the startup log showing only discord + telegram loaded but not WhatsApp.

cyan fjord
#

2026.4.26 isn't working for me also. very slow, once in a while comes to live and can chat with it in Telegram/Whatsap, but then goes silent..

twilit talonBOT
#

Yeah, this smells like a 2026.4.26 regression/slow-path rather than a normal channel config issue, especially since it affects Telegram + WhatsApp and Discord shows typing/restarts.

Can you run/paste this safe report?

openclaw --version
openclaw status --all
openclaw channels status --probe

Then, while reproducing one “goes silent” message:

openclaw logs --follow

Useful things to look for in the logs: model/provider timeout, repeated channel reconnects, Telegram polling stall, WhatsApp reconnect/flapping, or agent runs queued behind something else.

cyan fjord
#

thanks, but I already went through all that without much luck. decided to wait for next release hopefully it will be fixed

even basic commands such as openclaw status and openclaw logs can hang endlessly.