Skip to content

docs: Document runsync client timeout behavior#585

Draft
promptless[bot] wants to merge 1 commit intomainfrom
promptless/flash-runsync-timeout-behavior
Draft

docs: Document runsync client timeout behavior#585
promptless[bot] wants to merge 1 commit intomainfrom
promptless/flash-runsync-timeout-behavior

Conversation

@promptless
Copy link
Contributor

@promptless promptless bot commented Mar 23, 2026

Open this suggestion in Promptless to view citations and reasoning process

Documents that the Flash SDK's runsync() method uses execution_timeout_ms as the client-side HTTP timeout. This clarifies that long-running jobs (>60 seconds) need execution_timeout_ms configured to prevent client-side timeouts.

Trigger Events

runpod/flash PR #272: fix(serverless): use executionTimeoutMs as runsync client timeout

Promptless Research (5 files, 1 GitHub PR)

  • Reviewed runpod/flash#272
  • Viewed .long_term_context/product_knowledge/product_overview.md
  • Viewed .long_term_context/doc_workflow/client_instructions.md
  • Viewed flash/configuration/parameters.mdx
  • Viewed flash/apps/requests.mdx
  • Viewed .long_term_context/style/client_style_guide.md

Agent Response

I've created a documentation suggestion based on PR #272.

Summary of changes:

This PR fixes a bug where the Flash SDK's runsync() method had a hardcoded 60-second HTTP client timeout, regardless of the execution_timeout_ms setting. The fix makes runsync() respect the configured timeout, which is important for users running GPU inference jobs that take longer than a minute.

Documentation updates:

  1. flash/configuration/parameters.mdx — Added a Note to the execution_timeout_ms section explaining that the Flash SDK's runsync() method uses this value as the client-side HTTP timeout (defaulting to 60 seconds if unset).

  2. flash/apps/requests.mdx — Added a paragraph to the /runsync section explaining the timeout behavior and how to configure it for long-running jobs.

Suggestion: https://app.gopromptless.ai/suggestions/a8bd58f0-ea83-40ef-9e28-e4cef31b07e7


Tip: Point @Promptless at some of your docs debt and have it clean them up in the background 🧹

```

<Note>
The Flash SDK's `runsync()` method uses your `execution_timeout_ms` value as the client-side HTTP timeout. If set to a positive value, the SDK waits that duration for the job to complete. If unset or set to `0`, the SDK defaults to a 60-second timeout. For long-running inference jobs, set `execution_timeout_ms` to prevent premature timeouts.
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Citation: Bug fix PR that changed runsync() to use executionTimeoutMs / 1000 as client timeout instead of hardcoded 60s. Changes in src/runpod_flash/core/resources/serverless.py show the timeout derivation logic with guard for None and 0 values.
View source

}
```

The `/runsync` endpoint has a 60-second client-side timeout by default. If you've configured `execution_timeout_ms` on your endpoint, the client timeout uses that value instead. For jobs that take longer than 60 seconds, set `execution_timeout_ms` to prevent `/runsync` requests from timing out.
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Citation: PR #272 fixes ServerlessResource.runsync() to respect executionTimeoutMs for client-side HTTP timeout. The PR description explains the prior bug: 60s hardcoded timeout caused GPU inference jobs >1 minute to fail.
View source

@mintlify
Copy link

mintlify bot commented Mar 23, 2026

Preview deployment for your docs. Learn more about Mintlify Previews.

Project Status Preview Updated (UTC)
runpod-docs 🟢 Ready View Preview Mar 23, 2026, 5:41 PM

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants