openclaw - 💡(How to fix) Fix sessions_send should trigger fallback on 429/quota errors [1 participants]

If sessions_send receives a rate-limit / quota-style failure from the current model/provider, it should treat that as a fallback-eligible error and automatically try the next configured fallback.

• Make the final error explicit about which providers/models were attempted.

TL;DR

Modify the sessions_send function to treat HTTP 429 and provider quota exhaustion errors as fallback-eligible, allowing it to automatically try the next configured fallback provider.

Guidance

• Review the sessions_send function to identify where the fallback logic is currently implemented and modify it to catch and handle HTTP 429 and quota exhaustion errors.
• Update the error handling to distinguish between fallback-safe errors (e.g., 429, quota exhaustion) and non-fallback-safe errors, preserving the current behavior for the latter.
• Consider adding logging or error messaging to explicitly indicate which providers/models were attempted during the fallback process.
• Test the modified sessions_send function with a variety of error scenarios to ensure it correctly fails over to the next configured provider.

Example

No code snippet is provided as the issue does not include specific code details.

Notes

The solution assumes that the sessions_send function has a configurable fallback chain and that the current implementation has a mechanism for handling errors, but lacks the specific logic for treating HTTP 429 and quota exhaustion errors as fallback triggers.

Recommendation

Apply workaround: Modify the sessions_send function to treat HTTP 429 and provider quota exhaustion errors as fallback-eligible, as this will allow the function to automatically try the next configured fallback provider, improving the robustness of cross-session messaging.