View Single Post
Old 11-14-2013, 09:21 AM
ddcastillo ddcastillo is offline
Join Date: Nov 2013
Location: Coppell, Texas
Posts: 6
ddcastillo has 12 reputation points
Default AACC CachePing Functionality - CM (SIP) Versus CS1K (AML)

I wanted to cover this topic and see if it is helpful or interesting to anyone.

For many of our customers, whether you use a CM or a CS1K with your AACC HA solution, you may not understand the the design intent around the HA timeout value.

Avaya AACC Solutions:
Campus Mission Critical HA for SIP-enabled CC
Campus Hot-standby HA for AML-based CC
Campus Warm Standby HA

For sites using our AACC with CM, we offer an HA Solution that can monitor the network to a level of milliseconds. The design of this solution is to allow customers the ability to ensure their Call Center will function even with a very short network outage (as stated in milliseconds).

From our docs:
To support the Mission Critical HA switchover of contact center voice calls, SMMC must quickly determine if communication delays are due to normal network latency or if the delays are indicative of a network outage. SMMC balances the need for a rapid response to network outages with the need to avoid unnecessary switchovers. SMMC uses a network latency threshold (defined in milliseconds) to differentiate between normal network latency and a network outage.
• Network communication delays shorter than the threshold are treated as expected
• Network communication delays longer than the threshold are treated as potential network outages, and SMMC considers corrective action.

You must configure the High Availability SMMC “Network Timeout” threshold value high enough to be tolerant of normal network latency, but low enough to be responsive if a network failure occurs.

For sites using our AACC HA solution with a CS1K the timeout value is designed knowing that the site has a CS1K and therefore does not require the same level of HA. A CS1K allows the customer to provide ACD calls vai the CS1K. When designing this platform the designers have allowed for this and therefore the timeout value does not require a millisecond timeout value. AACC HA with CS1K allow for failovers within seconds and, it is up to the customer's IT team to determine what is normal network latency, and what is not, and set the HA timeout value appropriate so that a failover occurs at the time a true failover is appropriate.
Reply With Quote