How to Enable Redundancy with Microsoft Teams Call Recording
Note: This knowledgebase article applies only to Atmos’ Cloud-based (bot) recording of Microsoft Teams
Many Microsoft Teams call recording scenarios require redundancy to meet regulatory compliance, especially in financial, government, and medical markets. Atmos achieves redundancy using -two different methods outlined in this document. This article’s focus is redundancy, and it’s essential to understand the difference between redundancy, resilience, and high availability. Sometimes these terms are interchanged, but they have distinct meanings.
Redundancy refers to the purposeful duplication of a system’s components resulting in higher system dependability.
Resilience defines a given system’s power to recover from software and component faults and the degree of service persistence during fault events.
High availability (HA) is a trait of a system or network of systems designed to ensure a standard of performance, especially regarding system uptime, for above-average time periods.
Microsoft Teams Call Recording Redundancy In Atmos
There are two methods for enabling redundancy in Atmos’ Teams call recording scenarios.
N+1, also known as active standby, deploys a single bot into the users Azure space to capture Teams calls to a recording server. In N+1, there are two recording servers, a default server and a standby server. When the bot detects a fault in the default server, it will automatically switch to the standby server. CallCabinet initializes this default redundancy in Atmos for a Microsoft Teams recording system. The only potential fault in this scenario is if the bot itself fails. While this is very uncommon, the 2N redundancy method can provide a degree of resilience if it is a concern.
With 2N redundancy, two independent bots are deployed into the user’s space, each connected to a dedicated recording server. Both bots feed call audio to their servers directly without any switching or interaction in the event of a fault. The benefit of this method is independent redundancy, which is a somewhat more resilient redundancy scheme. Still, it’s important to mention that it requires twice the amount of storage space since two recordings are always made of every call.