Avaya MCC1 Maintenance Procedures page 204

Troubleshooting
Troubleshooting Multimedia Call Handling (MMCH)
Synchronization
Generally, the most common problem is a mismatch in synchronization sources between the endpoint and
the S8700 Media Server MMCH. This typically causes low-level (Px64) handshake problems that can
trigger the endpoint/MMCH to disconnect the call. The MCCH's timers are set to sufficiently high values
so that, normally, the endpoint will timeout and disconnect first. If installed in a customer network, it is a
good idea to perform an audit of the path synchronization is being supplied. If there are different clock
sources between endpoints and the S8700 Media Server MMCH, some problems are sure to occur. The
severity of these problems can range from a handshake failure every few seconds to one per day.
Depending on the type of endpoint, this can cause the endpoint to disconnect or just freeze video until the
main problem is resolved.
Specifically, PictureTel System 4000 endpoints seem to be the most sensitive to instability. The Avaya
Vistium disconnects fairly infrequently. The CLI Rembrandt II VP freezes video and waits for framing to
be recovered.
Network configuration concerns
with synchronization
When auditing a network for synchronization, avoid unnecessary hops. Thus, a switch providing star-
configuration synchronization is preferred over a daisy-chain configuration. Additionally, if there are
S8700 Media Server systems that have PNs, synchronization should be provided to subnodes from the
same PN through which the system receives its synchronization. Passing synchronization through a a
system's EI:
• Adds an unnecessary hop to the path
• Creates another potential point of failure
Expansion Interface duplication
If a customer's network uses PNs with duplicated EIs, scheduled switching of the EI links should be
disabled on the PBX by using change system-parameters maintenance. When scheduled maintenance
runs and switches the links, there is a brief corruption of the data path. If endpoints have active calls when
the switch occurs, this corruption of the data path causes Px64 handshake problems that lead to the
endpoints losing video source status and sometimes disconnecting as described above. Disabling the EI
switching is in the customer's best interest to prevent the disruption of the Px64 data stream. The
customer gets the same level of alarm indications and maintenance on the EI links, regardless of the
status of scheduled switching.
PRI D-channel backup
A somewhat unlikely source of call-stability problems can be incorrectly translated PRI D-channel
backup between two non-MCU systems. As an example:
• System A's DS1 1A10 is assigned as the primary source
• System B's corresponding DS1 is assigned as the secondary source
Then (when scheduled maintenance runs on either system with its backup D channel active), an audit
disconnects some calls using the link. This problem is corrected by assigning matching
primary/secondary D channels for the two systems.
Processor duplication in the system
Do not enable the PI link switch during scheduled maintenance. This can cause link-stability problems
on the Accunet Bandwidth Controller (ABC).
204 Maintenance Procedures
December 2003