[GH-ISSUE #24042] issue: [BUG] 100% CPU after cancel / streaming completion (anyio cancellation loop)

Originally created by @tbc3697 on GitHub (Apr 23, 2026). Original GitHub issue: https://github.com/open-webui/open-webui/issues/24042 ### Check Existing Issues - [x] I have searched for any existing and/or related issues. - [x] I have searched for any existing and/or related discussions. - [x] I have also searched in the CLOSED issues AND CLOSED discussions and found no related items (your issue might already be addressed on the development branch!). - [x] I am using the latest version of Open WebUI. ### Installation Method Docker ### Open WebUI Version v0.9.1 ### Ollama Version (if applicable) _No response_ ### Operating System mac os 26.1, docker ### Browser (if applicable) _No response_ ### Confirmation - [x] I have read and followed all instructions in `README.md`. - [x] I am using the latest version of **both** Open WebUI and Ollama. - [x] I have included the browser console logs. - [x] I have included the Docker container logs. - [x] I have **provided every relevant configuration, setting, and environment variable used in my setup.** - [x] I have clearly **listed every relevant configuration, custom setting, environment variable, and command-line option that influences my setup** (such as Docker Compose overrides, .env values, browser settings, authentication configurations, etc). - [x] I have documented **step-by-step reproduction instructions that are precise, sequential, and leave nothing to interpretation**. My steps: - Start with the initial platform/version/OS and dependencies used, - Specify exact install/launch/configure commands, - List URLs visited, user input (incl. example values/emails/passwords if needed), - Describe all options and toggles enabled or changed, - Include any files or environmental changes, - Identify the expected and actual result at each stage, - Ensure any reasonably skilled user can follow and hit the same issue. ### Expected Behavior Cancelling a completed or nearly completed request should: * Gracefully exit * Not leave any running async loops * CPU should return to idle ### Actual Behavior * CPU usage spikes to ~100% * Process does not recover * No additional requests are needed to trigger the issue ### Steps to Reproduce Case 1 (100% reproducible) * Send a request (streaming enabled) * Wait until the model finishes generating completely * Click "Stop" / "End" button * Observe CPU usage 👉 Result: CPU jumps to ~100% and stays there indefinitely Case 2 (partially reproducible) * Send a request (streaming enabled) * Wait until generation completes * Perform one of the following: * Refresh page * Close tab * Logout 👉 Result: CPU spike occurs sometimes (not 100%) ### Logs & Screenshots No ### Additional Information Downgrading anyio reduces likelihood but does not fully eliminate the issue * Issue is much easier to reproduce with: * streaming enabled * tool / multi-step execution (native mode) * Explicit cancel (UI "Stop") is 100% reproducible * Connection close (refresh/tab close) is timing-dependent

GiteaMirror commented

2026-04-25 09:52:50 -05:00

Owner

Originally created by @tbc3697 on GitHub (Apr 23, 2026).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/24042

Check Existing Issues

I have searched for any existing and/or related issues.
I have searched for any existing and/or related discussions.
I have also searched in the CLOSED issues AND CLOSED discussions and found no related items (your issue might already be addressed on the development branch!).
I am using the latest version of Open WebUI.

Installation Method

Docker

Open WebUI Version

v0.9.1

Ollama Version (if applicable)

No response

Operating System

mac os 26.1, docker

Browser (if applicable)

Confirmation

I have read and followed all instructions in README.md.
I am using the latest version of both Open WebUI and Ollama.
I have included the browser console logs.
I have included the Docker container logs.
I have provided every relevant configuration, setting, and environment variable used in my setup.
I have clearly listed every relevant configuration, custom setting, environment variable, and command-line option that influences my setup (such as Docker Compose overrides, .env values, browser settings, authentication configurations, etc).
I have documented step-by-step reproduction instructions that are precise, sequential, and leave nothing to interpretation. My steps:
Start with the initial platform/version/OS and dependencies used,
Specify exact install/launch/configure commands,
List URLs visited, user input (incl. example values/emails/passwords if needed),
Describe all options and toggles enabled or changed,
Include any files or environmental changes,
Identify the expected and actual result at each stage,
Ensure any reasonably skilled user can follow and hit the same issue.

Expected Behavior

Cancelling a completed or nearly completed request should:

Gracefully exit
Not leave any running async loops
CPU should return to idle

Actual Behavior

CPU usage spikes to ~100%
Process does not recover
No additional requests are needed to trigger the issue

Steps to Reproduce

Case 1 (100% reproducible)

Send a request (streaming enabled)
Wait until the model finishes generating completely
Click "Stop" / "End" button
Observe CPU usage

👉 Result: CPU jumps to ~100% and stays there indefinitely

Case 2 (partially reproducible)

Send a request (streaming enabled)
Wait until generation completes
Perform one of the following:
Refresh page
Close tab
Logout

👉 Result: CPU spike occurs sometimes (not 100%)

Logs & Screenshots

Additional Information

Downgrading anyio reduces likelihood but does not fully eliminate the issue

Issue is much easier to reproduce with:
streaming enabled
- tool / multi-step execution (native mode)
- Explicit cancel (UI "Stop") is 100% reproducible
Connection close (refresh/tab close) is timing-dependent

GiteaMirror added the bug label 2026-04-25 09:52:50 -05:00

[GH-ISSUE #24042] issue: [BUG] 100% CPU after cancel / streaming completion (anyio cancellation loop) #35694