mirror of
https://github.com/openai/codex.git
synced 2026-11-15 21:50:46 +00:00
Fall back to HTTP immediately for oversized WebSocket requests (#60840)
## Why A WebSocket close with code `1009` before any text arrives indicates that the request exceeds the peer's message size limit. Retrying the request over another WebSocket does not address that limit. ## What changed - Classify these early size rejections as requiring HTTP fallback and switch transports without exhausting WebSocket retries, including during prewarming. - Preserve ordinary stream retry behavior for size closes after any text frame and for other close codes. ## Testing Add integration coverage for early size rejections during generation and prewarming, size closes after metadata, rate limits, or response creation, and ordinary closes. GitOrigin-RevId: 4550d289bb414fec1faae2a195bcc635d394117b
This commit is contained in:
@@ -2571,6 +2571,10 @@ impl ModelClientSession {
|
||||
while let Some(event) = stream.next().await {
|
||||
match event {
|
||||
Ok(ResponseEvent::Completed { .. }) => break,
|
||||
Err(err) if err.requires_http_fallback() => {
|
||||
self.try_switch_fallback_transport(session_telemetry, model_info);
|
||||
break;
|
||||
}
|
||||
Err(err) => return Err(err),
|
||||
_ => {}
|
||||
}
|
||||
|
||||
@@ -118,12 +118,13 @@ pub(crate) async fn handle_response_stream_error(
|
||||
return Ok(());
|
||||
}
|
||||
|
||||
if retry_state.retries >= max_retries
|
||||
let requires_http_fallback = err.requires_http_fallback();
|
||||
let switched = (requires_http_fallback || retry_state.retries >= max_retries)
|
||||
&& client_session.try_switch_fallback_transport(
|
||||
&turn_context.session_telemetry,
|
||||
turn_context.model_info(),
|
||||
)
|
||||
{
|
||||
);
|
||||
if switched {
|
||||
// Changing transport must not bypass the server's retry deadline.
|
||||
if let Some(retry_after) = retry_after {
|
||||
tokio::time::sleep_until(retry_after.deadline()).await;
|
||||
@@ -135,6 +136,9 @@ pub(crate) async fn handle_response_stream_error(
|
||||
}),
|
||||
)
|
||||
.await;
|
||||
}
|
||||
// Another request can already have switched this session to HTTP.
|
||||
if switched || requires_http_fallback {
|
||||
retry_state.retries = 0;
|
||||
return Ok(());
|
||||
}
|
||||
|
||||
@@ -3158,7 +3158,10 @@ async fn try_run_sampling_request(
|
||||
.await;
|
||||
should_emit_token_count |= incomplete.token_usage.is_some();
|
||||
}
|
||||
if *draining_for_mailbox && matches!(err.details(), CodexErrorDetails::Stream(_)) {
|
||||
if *draining_for_mailbox
|
||||
&& (matches!(err.details(), CodexErrorDetails::Stream(_))
|
||||
|| err.requires_http_fallback())
|
||||
{
|
||||
// Mail already requires a follow-up, even if the producer exited before
|
||||
// receiving the interrupt. Do not retry without that pending input.
|
||||
break Ok(SamplingRequestResult {
|
||||
|
||||
Reference in New Issue
Block a user