High Availability & Disaster Recovery

overview this document explains ameyo’s high availability (ha) and disaster recovery (dr) strategies for maintaining continuous system operations it covers ha failover methods and dr processes that ensure minimal downtime and quick restoration during failures 🖥️ high availability (ha) high availability refers to a system's ability to remain operational and available despite the failure of one or more of its components ha provides service redundancy it relies on redundant components and a "switchover" process (either automatic or manual) to a standby component upon failure high availability for software high availability software refers to the use of software to ensure that systems are running (available) most of the time high availability is a characteristic of a system and is defined as the percentage of time that the system is functioning ha cluster a high availability cluster is a group of hosts that act like a single system and provide continuous uptime high availability clusters are often used for load balancing, backup and failover purposes clustering clustering is the most common technique to achieve high availability for any services by introducing redundancy in software, hardware and data in a failure the clustering software immediately start the application on the standby system without requiring administrative intervention depending on the type of redundancy in software to be provided for high availability, clusters can be configured in any of the following configurations 1\ automatic failover (hot failover) this method, also known as hot failover , is designed for minimal to zero disruption as it does not require manual intervention the transition to the redundant component is transparent, allowing users to continue work without interruption ameyo's hot failover solution promises an impressive uptime of 99 95% example application server failure in a hot failover setup, agents are connected to the ameyo application via a virtual ip if the main application server goes down, the redundant server automatically takes over the agent might experience a brief connection error (1 2 minutes maximum) before being automatically reconnected critically, existing calls continue without any manual intervention in ameyo, there are three ways of implementing hot failover depending on the setup size and requirement of organization 1+1 redundancy 1+1 redundancy can be provided in those cases where organizations requires separate server for managing high availability for all services (app + database + call) under the solution n+1 redundancy n+1 redundancy can be available in those cases where organizations is asking for a single server for managing redundancy for all services (app + database + call) running on different servers n+n redundancy as the name suggesting, n+n redundancy is something where multiple redundancy servers are involved for managing ameyo services on multiple primary servers 2\ manual failover (cold failover) this method, known as cold failover , involves replicating ameyo services on a separate standby server this maintains a copy of the services that can be used during a fault this approach is suitable for organizations with smaller contact centers that are not running mission critical processes and can absorb a brief, planned downtime ameyo's cold failover solution promises 99 5% uptime main server is active how it works unlike hot failover, this process requires manual intervention from an it team when the main server fails, a technician must disconnect the main server from the lan configure the standby server with the same ip address as the failed server manually start all ameyo services on the (now active) standby server once these steps are complete, operations can resume standby server is activated manually when main server is down] 3\ load sharing (active active) ameyo also implements an active active methodology for call services, which acts as a form of load sharing in this configuration, call managers run on both the main and redundant servers simultaneously calls are distributed across both active call managers sip proxy services allow contact center users to register on both managers, ensuring high availability if one call manager fails, the calls on that specific server are disconnected however, the secondary call manager is still functional and immediately begins handling the full load of all new calls 🧯 disaster recovery (dr) while ha handles component failure (like a single server), disaster recovery handles large scale disasters (natural or man made) dr provides data recovery it is a formal, documented plan for restoring a system to an acceptable state after a disaster, such as the loss of an entire building or data center in a dr event, all call center operations are diverted to a separate, pre determined dr site ameyo's dr methodology the ameyo suite provides a full fledged dr solution using real time database replication the main (central location) and dr location are kept continuously updated over a stable connection, such as mpls activating the dr site (standard plan of action) when a disaster occurs and the main site goes down, the dr site can be made live with only a few operational and solution level changes the high level steps to activate the ameyo business continuity plan (bcp) are as follows call centre movement the organization moves call center staff to the dr site or another location where the ameyo dr setup is accessible access dr site setup using a cli (command line interface) terminal, the it team accesses the dr site configuration db mode change the dr site's database server, which is normally in read mode, must be changed to write mode start services the required ameyo licensing and application services are started in the correct order agent login agents are required to log in to the ameyo application by changing the ip address in their web url and within their softphone configuration