Teams Direct routing media bypass mode ICE controlled
I've a strange scenario when I enabled media bypass on my SBC and thought to check with everyone here first before contacting MS. Outbound calls from Teams to PSTN works with ICE being CONTROLLING from Teams end (FULL ICE) and once the call is established Teams starts nominating the ICE candidates via a re-invite and it's handled correctly and audio is flowing in both directions. Inbound calls from PSTN to Teams is having a strange problem from Teams endpoint as the SBC should be ICE-LITE mode (controlled) and not doing any connectivity checks , just responding to them. and SDP being sent looks perfect like the following a=rtcp-mux a=crypto:1 AES_CM_128_HMAC_SHA1_80 inline:/MdAvE9GIDAg7H6ydytIrVyKbgrirK614YLxHj/F|2^31 a=ice-lite a=ice-ufrag:wIoWCV4A a=ice-pwd:fD9WHcj3cCbNPcxDjU9Cp0AUEt a=candidate:CeTxOSQoHkYUy9st 1 UDP 2130706431 xxx.xxx.xxx.xxx 30036 typ host however Teams endpoints are sending their STUN requests as CONTROLLED !! while it should be CONTROLLING. yes audio is flowing fine but am not getting the nomination via the re-invite which of course is wrong any suggestions ?
Media Optimization for Direct Routing: Failover of ProxySBC for PSTNGateway
We are currently building Direct Routing voice infra for MS Teams with Media Optimization and Always Bypass mode using Audiocodes SBCs. Aim is to have two centralized ProxySBCs, primary and standby, shared over many PSTNGateways. When we define PSTN Gateway and Virtual Network Topology in Teams we can only define a single ProxySBC parameter for each PSTNGateway: PS C:\> Set-CsOnlinePSTNGateway -Identity <Identity> -GatewaySiteID <site ID> -MediaBypass <true/false> -BypassMode <Always/OnlyForLocalUsers> -ProxySBC <proxy SBC FQDN or $null> We would like to define geographical failover strategy for Calls to use standby ProxySBC. Are you aware about any possible option to achieve automatic ProxySBC failover? Options that comes to my mind, we could: Duplicate PSTNGateways once with Primary ProxySBC and second with Standby ProxySBC and play with VoicePolicies, but that would involve 2 different FQDNs and 2 Certificates for 1 same thing. Do Manual failover to Standby ProxySBC with a script but that misses automation. Build HA Pair but that's not available in two geographical different locations in our environment. Any option I overlooked how Microsoft plan ProxySBC failover? Any comments greatly appreciated.
