Repository navigation
Initialization of signaling connection too slow in some cases #1403
Description
We are using the livekit sdk (through the python sdk) and in some ocassions (~0.5%) we get sequences of livekit logs like these:
2026-09-03 10:27:35.395 | livekit_api::signal_client::signal_stream:106:livekit_api::signal_client::signal_stream - connecting to wss://xxx.livekit.cloud/rtc?...
2026-09-03 10:28:07.034 | rustls::anchors:150:rustls::anchors - add_parsable_certificates processed 146 valid and 0 invalid certs
2026-09-03 10:28:07.035 | tokio_tungstenite::tls::encryption::rustls:103:tokio_tungstenite::tls::encryption::rustls - Added 146/146 native root certificates (ignored 0)
2026-09-03 10:28:07.035 | rustls::client::hs:73:rustls::client::hs - No cached session for DnsName("xxx.livekit.cloud")
2026-09-03 10:28:07.035 | rustls::client::hs:132:rustls::client::hs - Not resuming any session
2026-09-03 10:28:07.060 | rustls::client::hs:615:rustls::client::hs - Using ciphersuite TLS13_AES_128_GCM_SHA256
2026-09-03 10:28:07.061 | rustls::client::tls13:142:rustls::client::tls13 - Not resuming
2026-09-03 10:28:07.061 | rustls::client::tls13:381:rustls::client::tls13 - TLS1.3 encrypted extensions: []
2026-09-03 10:28:07.061 | rustls::client::hs:472:rustls::client::hs - ALPN protocol is None
2026-09-03 10:28:07.156 | tungstenite::handshake::client:95:tungstenite::handshake::client - Client handshake done.
As you can see it took many seconds (in this case 30s but i've seen a few even higher cases) to initialize de rustls stuff. Is it possible to optimize that or do some type of prewarming in advance to avoid that situation?
Activity