Cisco Cisco TelePresence Video Communication Server Expressway 릴리즈 노트
Resolved caveats
Cisco TelePresence Video Communication Server X7.2.4 Software Release Notes
Page 29 of 50
Identifier
Summary
Workaround: None.
Additional Information: If VCS requests a SIP call to be cleared (BYE) but never gets an
ACK, it should clear the call, including the H.323 leg if a BYE is received on the SIP side.
Additional Information: If VCS requests a SIP call to be cleared (BYE) but never gets an
ACK, it should clear the call, including the H.323 leg if a BYE is received on the SIP side.
CSCtw82611
Symptom: Presentation is shown as a black screen to all participants in the MCU conference.
Conditions: C-series codec (C40 and C60 tested so far) running TC4.2.0 calls (SIP) into an
MCU conference (4.2(1.43)) (H323 only) with another movi participant running (4.2) (Calls are
interworked by the VCS). The C-series codec shares presentation to the conference. Then the
C-series uses its multi-site to patch in an audio-only participant (This part does not need to be
interworked to recreate the bug).
After 15 minutes of all participants being in the conference, if the C-series endpoint stops
presentation and starts it again, members of the conference see a black screen instead of the
presentation.
Workaround: None currently known.
Conditions: C-series codec (C40 and C60 tested so far) running TC4.2.0 calls (SIP) into an
MCU conference (4.2(1.43)) (H323 only) with another movi participant running (4.2) (Calls are
interworked by the VCS). The C-series codec shares presentation to the conference. Then the
C-series uses its multi-site to patch in an audio-only participant (This part does not need to be
interworked to recreate the bug).
After 15 minutes of all participants being in the conference, if the C-series endpoint stops
presentation and starts it again, members of the conference see a black screen instead of the
presentation.
Workaround: None currently known.
CSCtx24759
Symptoms: The VCS suffers high CPU load when Presence is in use, leading to reduced
responsiveness.
Conditions: In versions prior to X5.2, the default value for retries in the Presence User Agent
was 5 seconds. This was changed to 1800 seconds, but existing configuration was not updated
during upgrades. This causes unnecessary retries in situations where some domains on a VCS
do not have a presence server.
Workaround: Set the value to the new default with the following CLI command: xConfiguration
Applications Presence User Agent RetryDelta: 1800
responsiveness.
Conditions: In versions prior to X5.2, the default value for retries in the Presence User Agent
was 5 seconds. This was changed to 1800 seconds, but existing configuration was not updated
during upgrades. This causes unnecessary retries in situations where some domains on a VCS
do not have a presence server.
Workaround: Set the value to the new default with the following CLI command: xConfiguration
Applications Presence User Agent RetryDelta: 1800
CSCtx33677
Symptom: An interface on the VCS appears to be unreachable to the network.
Conditions: VCS X7 or later, with the dual nic key and both interfaces on the same subnet.
Workaround: Move the interfaces to different subnets.
Resolution: Explicitly stated in the X7.1 help/admin guide that "The LAN 1 and LAN 2
interfaces must be on different, non-overlapping subnets."
Conditions: VCS X7 or later, with the dual nic key and both interfaces on the same subnet.
Workaround: Move the interfaces to different subnets.
Resolution: Explicitly stated in the X7.1 help/admin guide that "The LAN 1 and LAN 2
interfaces must be on different, non-overlapping subnets."
CSCty01037
Symptoms: Success of audio-only H.323 calls being invited into a Multiway conference by a
SIP endpoint via a VCS may not be correctly reported to the SIP endpoint. This leads to the
SIP endpoint (e.g. C40) failing on Multiway Join.
Conditions: This requires the VCS to be acting as an H.323/SIP interworking gateway, with a
neighbored MCU. The H.323 participants are audio-only, and a SIP participant (such as a C40)
is the Multiway initiator. On initiating Multiway Join, the feedback of success is not correctly
returned to the SIP endpoint.
Workaround: This problem will not occur on H.323 only.
SIP endpoint via a VCS may not be correctly reported to the SIP endpoint. This leads to the
SIP endpoint (e.g. C40) failing on Multiway Join.
Conditions: This requires the VCS to be acting as an H.323/SIP interworking gateway, with a
neighbored MCU. The H.323 participants are audio-only, and a SIP participant (such as a C40)
is the Multiway initiator. On initiating Multiway Join, the feedback of success is not correctly
returned to the SIP endpoint.
Workaround: This problem will not occur on H.323 only.
Resolved in X7.0.3
Identifier
Summary
CSCtu06577
VCS may crash and report a SIGSEGV fault:
Symptoms: Cisco VCS may crash, and an alarm be raised stating that an unexpected
software error was detected in app with a SIGSEGV fault. A crash report will also be generated
and when this is analyzed a call to sha1_block_data_order() in libcrypto is seen in the trace
back.
Conditions: This was on a Cisco VCS release X7.0.1
Workaround: There is no workaround at this time.
Symptoms: Cisco VCS may crash, and an alarm be raised stating that an unexpected
software error was detected in app with a SIGSEGV fault. A crash report will also be generated
and when this is analyzed a call to sha1_block_data_order() in libcrypto is seen in the trace
back.
Conditions: This was on a Cisco VCS release X7.0.1
Workaround: There is no workaround at this time.
CSCtt94053
No presence PUBLISH generated by PUA in SIP->H323 interworked call in VCS cluster:
Symptoms: sometimes in-call presence is not published by PUA for H.323 calls.
Conditions: H.323 endpoint must register to a cluster peer and the call must come in through
another peer in the cluster. (If the call arrives on the same peer to which the endpoint is
registered, the in-call presence will be fine.)
Workaround: dual register endpoints as SIP and H.323.
Symptoms: sometimes in-call presence is not published by PUA for H.323 calls.
Conditions: H.323 endpoint must register to a cluster peer and the call must come in through
another peer in the cluster. (If the call arrives on the same peer to which the endpoint is
registered, the in-call presence will be fine.)
Workaround: dual register endpoints as SIP and H.323.
CSCtw61291
The VCS fails to listen on call signaling ports:
Symptoms: The VCS fails to listen on call signaling ports after a reboot / restart / upgrade.
Symptoms: The VCS fails to listen on call signaling ports after a reboot / restart / upgrade.