Description:

When using a local Ollama model and the LLM request times out (e.g., due to a slow response from a large model), OpenClaw puts the Ollama provider into "cooldown" as if it were a rate limit from a paid API. This makes no sense for local models that have no rate limits.
Steps to reproduce:

Configure Ollama as a provider with a local model (e.g., qwen3-coder-next)
Send a complex request that takes longer than the timeout threshold
Observe error: Provider ollama is in cooldown (all profiles unavailable) (rate_limit)
All subsequent requests fail even though Ollama is running and accessible

Expected behavior:
Timeouts from local providers should not trigger rate-limit cooldown. The request should simply retry or fail gracefully without locking out the provider.

Actual behavior:
The provider is marked as unavailable in ~/.openclaw/agents/main/agent/auth-profiles.json with disabledUntil timestamp. Manual editing of this file is required to restore functionality.
Workaround:
Edit ~/.openclaw/agents/main/agent/auth-profiles.json and delete the usageStats section, then restart the gateway.

Environment:
Windows 11 Pro
Docker for Windows
WSL2
OpenClaw version: 2026.2.4
Ollama version: 0.15.6
Running in Docker on WSL2
Model: qwen3-coder-next (80B MoE)

Logs or screenshots

andrew@NucBoxEVO-X2:~/openclaw$ docker compose logs openclaw-gateway | tail -50
WARN[0000] The "CLAUDE_WEB_COOKIE" variable is not set. Defaulting to a blank string.
WARN[0000] The "CLAUDE_AI_SESSION_KEY" variable is not set. Defaulting to a blank string.
WARN[0000] The "CLAUDE_WEB_SESSION_KEY" variable is not set. Defaulting to a blank string.
WARN[0000] The "CLAUDE_WEB_COOKIE" variable is not set. Defaulting to a blank string.
WARN[0000] The "CLAUDE_AI_SESSION_KEY" variable is not set. Defaulting to a blank string.
WARN[0000] The "CLAUDE_WEB_SESSION_KEY" variable is not set. Defaulting to a blank string.
openclaw-gateway-1 | 2026-02-10T10:26:06.698Z [gateway] log file: /tmp/openclaw/openclaw-2026-02-10.log
openclaw-gateway-1 | 2026-02-10T10:26:06.710Z [browser/service] Browser control service ready (profiles=2)
openclaw-gateway-1 | 2026-02-10T10:26:06.893Z [gateway] update available (latest): v2026.2.9 (current v2026.2.4). Run: openclaw update
openclaw-gateway-1 | 2026-02-10T10:26:08.659Z [telegram] [default] starting provider (@ElliotOpenClawBot)
openclaw-gateway-1 | 2026-02-10T10:26:08.674Z [telegram] autoSelectFamily=false (default-node22)
openclaw-gateway-1 | 2026-02-10T10:26:09.524Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:26:10.379Z [ws] webchat connected conn=3942d54d-d817-46f0-ac3b-9a19dfc9796f remote=172.18.0.1 client=openclaw-control-ui webchat vdev
openclaw-gateway-1 | 2026-02-10T10:26:11.099Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:26:11.248Z [agent/embedded] Removed orphaned user message to prevent consecutive user turns. runId=28f1c057-b019-437a-b221-e8a9f230f9e9 sessionId=06ba8853-3cb1-44cc-8dc5-c36e04afe115
openclaw-gateway-1 | 2026-02-10T10:27:22.525Z [ws] ⇄ res ✓ chat.history 55ms conn=3942d54d…796f id=9beef74b…74af
openclaw-gateway-1 | 2026-02-10T10:27:31.010Z [ws] ⇄ res ✓ chat.history 52ms conn=3942d54d…796f id=14a5b28d…983d
openclaw-gateway-1 | 2026-02-10T10:28:00.077Z [ws] ⇄ res ✓ config.schema 111ms conn=3942d54d…796f id=7ae31d94…bebb
openclaw-gateway-1 | 2026-02-10T10:28:06.617Z typing TTL reached (2m); stopping typing indicator
openclaw-gateway-1 | 2026-02-10T10:28:19.205Z [ws] ⇄ res ✓ chat.history 63ms conn=3942d54d…796f id=e5deb551…6722
openclaw-gateway-1 | 2026-02-10T10:29:01.018Z [ws] ⇄ res ✓ chat.history 51ms conn=3942d54d…796f id=a0d43e11…1aea
openclaw-gateway-1 | 2026-02-10T10:29:02.443Z [ws] ⇄ res ✓ chat.history 56ms conn=3942d54d…796f id=25941893…515b
openclaw-gateway-1 | 2026-02-10T10:35:54.831Z [agent/embedded] embedded run timeout: runId=28f1c057-b019-437a-b221-e8a9f230f9e9 sessionId=06ba8853-3cb1-44cc-8dc5-c36e04afe115 timeoutMs=600000
openclaw-gateway-1 | 2026-02-10T10:35:54.847Z [agent/embedded] Profile ollama:default timed out (possible rate limit). Trying next account...
openclaw-gateway-1 | 2026-02-10T10:35:54.848Z [diagnostic] lane task error: lane=main durationMs=583751 error="FailoverError: LLM request timed out."
openclaw-gateway-1 | 2026-02-10T10:35:54.849Z [diagnostic] lane task error: lane=session:agent:main:main durationMs=583754 error="FailoverError: LLM request timed out."
openclaw-gateway-1 | 2026-02-10T10:35:54.854Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:36:09.445Z [diagnostic] lane task error: lane=main durationMs=14592 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:09.446Z [diagnostic] lane task error: lane=session:agent:main:main durationMs=14594 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:09.454Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:36:09.472Z [diagnostic] lane task error: lane=main durationMs=22 error="FailoverError: No available auth profile for ollama (all in cooldown or unavailable)."
openclaw-gateway-1 | 2026-02-10T10:36:09.473Z [diagnostic] lane task error: lane=session:agent:main:main durationMs=25 error="FailoverError: No available auth profile for ollama (all in cooldown or unavailable)."
openclaw-gateway-1 | 2026-02-10T10:36:09.477Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:36:09.494Z [diagnostic] lane task error: lane=main durationMs=18 error="FailoverError: No available auth profile for ollama (all in cooldown or unavailable)."
openclaw-gateway-1 | 2026-02-10T10:36:09.495Z [diagnostic] lane task error: lane=session:agent:main:main durationMs=20 error="FailoverError: No available auth profile for ollama (all in cooldown or unavailable)."
openclaw-gateway-1 | 2026-02-10T10:36:09.496Z Embedded agent failed before reply: All models failed (5): ollama/qwen3-coder-next: LLM request timed out. (unknown) | anthropic/claude-opus-4-5: Provider anthropic is in cooldown (all profiles unavailable) (rate_limit) | openai/gpt-4o: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors. (rate_limit) | ollama/qwen3:30b: No available auth profile for ollama (all in cooldown or unavailable). (rate_limit) | ollama/qwen3-coder-next:latest: No available auth profile for ollama (all in cooldown or unavailable). (rate_limit)
openclaw-gateway-1 | 2026-02-10T10:36:09.507Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:36:17.200Z [diagnostic] lane task error: lane=main durationMs=7700 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:17.200Z [diagnostic] lane task error: lane=session:agent:main:main durationMs=7702 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:17.201Z Followup agent failed before reply: All models failed (5): ollama/qwen3-coder-next:latest: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | anthropic/claude-opus-4-5: Provider anthropic is in cooldown (all profiles unavailable) (rate_limit) | openai/gpt-4o: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors. (rate_limit) | ollama/qwen3:30b: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | ollama/qwen3-coder-next: Provider ollama is in cooldown (all profiles unavailable) (rate_limit)
openclaw-gateway-1 | 2026-02-10T10:36:17.205Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:36:26.830Z [diagnostic] lane task error: lane=main durationMs=9626 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:26.830Z [diagnostic] lane task error: lane=session:agent:main:main durationMs=9628 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:26.831Z Followup agent failed before reply: All models failed (5): ollama/qwen3-coder-next:latest: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | anthropic/claude-opus-4-5: Provider anthropic is in cooldown (all profiles unavailable) (rate_limit) | openai/gpt-4o: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors. (rate_limit) | ollama/qwen3:30b: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | ollama/qwen3-coder-next: Provider ollama is in cooldown (all profiles unavailable) (rate_limit)
openclaw-gateway-1 | 2026-02-10T10:36:26.834Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:36:34.945Z [diagnostic] lane task error: lane=main durationMs=8111 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:34.946Z [diagnostic] lane task error: lane=session:agent:main:main durationMs=8113 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:34.946Z Followup agent failed before reply: All models failed (5): ollama/qwen3-coder-next:latest: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | anthropic/claude-opus-4-5: Provider anthropic is in cooldown (all profiles unavailable) (rate_limit) | openai/gpt-4o: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors. (rate_limit) | ollama/qwen3:30b: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | ollama/qwen3-coder-next: Provider ollama is in cooldown (all profiles unavailable) (rate_limit)
openclaw-gateway-1 | 2026-02-10T10:36:34.949Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:36:41.334Z [diagnostic] lane task error: lane=main durationMs=6385 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:41.335Z [diagnostic] lane task error: lane=session:agent:main:main durationMs=6387 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:41.335Z Followup agent failed before reply: All models failed (5): ollama/qwen3-coder-next:latest: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | anthropic/claude-opus-4-5: Provider anthropic is in cooldown (all profiles unavailable) (rate_limit) | openai/gpt-4o: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors. (rate_limit) | ollama/qwen3:30b: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | ollama/qwen3-coder-next: Provider ollama is in cooldown (all profiles unavailable) (rate_limit)
openclaw-gateway-1 | 2026-02-10T10:36:41.342Z Failed to discover Ollama models: TypeError: fetch failed
openclaw-gateway-1 | 2026-02-10T10:36:49.375Z [diagnostic] lane task error: lane=main durationMs=8037 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:49.375Z [diagnostic] lane task error: lane=session:agent:main:main durationMs=8038 error="FailoverError: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors."
openclaw-gateway-1 | 2026-02-10T10:36:49.376Z Followup agent failed before reply: All models failed (5): ollama/qwen3-coder-next:latest: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | anthropic/claude-opus-4-5: Provider anthropic is in cooldown (all profiles unavailable) (rate_limit) | openai/gpt-4o: You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors. (rate_limit) | ollama/qwen3:30b: Provider ollama is in cooldown (all profiles unavailable) (rate_limit) | ollama/qwen3-coder-next: Provider ollama is in cooldown (all profiles unavailable) (rate_limit)