litellm

mirror of https://github.com/BerriAI/litellm.git synced 2026-09-11 22:51:28 +00:00

History

Mateo 9271133beb fix(realtime): bound Vertex credential resolution and make realtime failures loud A /v1/realtime connection to a Vertex AI Live model accepted the WebSocket upgrade and then went silent: a stalled Google OAuth token fetch blocked the handler before any session event, and the eventual failure closed the socket with a bare 1011 and no error event, so callers saw an open socket, no frames, and no reason. Bound the pre-session token fetch with REALTIME_CREDENTIAL_RESOLUTION_TIMEOUT_SECONDS (20s default) and, on any realtime failure, send an OpenAI-style error event before closing with a reason that names the failure. Close reasons are truncated by bytes, not characters, since an over-long reason makes the close frame itself fail.	2026-08-20 02:02:00 -07:00
..
test_realtime_webrtc_endpoints.py	fix(realtime): bound Vertex credential resolution and make realtime failures loud	2026-08-20 02:02:00 -07:00

Mateo 9271133beb fix(realtime): bound Vertex credential resolution and make realtime failures loud

A /v1/realtime connection to a Vertex AI Live model accepted the WebSocket
upgrade and then went silent: a stalled Google OAuth token fetch blocked the
handler before any session event, and the eventual failure closed the socket
with a bare 1011 and no error event, so callers saw an open socket, no frames,
and no reason.

Bound the pre-session token fetch with
REALTIME_CREDENTIAL_RESOLUTION_TIMEOUT_SECONDS (20s default) and, on any
realtime failure, send an OpenAI-style error event before closing with a reason
that names the failure. Close reasons are truncated by bytes, not characters,
since an over-long reason makes the close frame itself fail.

2026-08-20 02:02:00 -07:00

test_realtime_webrtc_endpoints.py

fix(realtime): bound Vertex credential resolution and make realtime failures loud

2026-08-20 02:02:00 -07:00