High Availability and Failover Solutions

Symlabs' Virtual Directory Server makes your directory infrastructure highly-available. An extensive set of algorithms has been elaborated to monitor servers and route around any potential problem. Virtual Directory Server is currently running in mission-critical environments at organizations all over the world. Many of these deployments involve very high performance demands and enormous numbers of entries numbering to the hundreds of millions.

Solution

Automatic Failover

Directories are typically deployed with replication enabled. This makes perfect sense, because when one directory server goes down for any reason, another server is standing by to take the load. However, applications are usually not capable of using a stand-by server when the primary server is down.

Virtual Directory Server creates a fail-safe environment by routing around unavailable directory servers. This happens in a transparent way, where operations that have not yet been responded to are automatically re-sent to a backup server when a primary server goes down.

Multiple Failover Algorithms for full Flexibility

When implementing high availability for your directory infrastructure, one size does not fit all requirements. Depending on the replication strategy used, some failover mechanisms work well, while others can be counter-productive. The failover algorithm must provide high availability without compromising data integrity.

That is why Symlabs Virtual Directory Server comes with a variety of different failover algorithms designed for the different scenarios in which directory infrastructure can be deployed. For example, for multi-master replication, the affinity algorithm is advisable, because it avoids requests sent to both masters at the same time, therefore avoiding potential replication collisions. The multisite algorithm is ideal for organizations with disaster recovery sites across multiple sites, allowing a weighted prioritization of connections on a per-site basis. There are several other algorithms to choose from, depending on the replication strategy used.

Health Monitoring

Virtual Directory Server supports health monitoring by periodically checking each server's state. Whenever servers go down, immediate action is taken and traffic is rerouted accordingly. When servers come back up, depending on the failover algorithm chosen, connections are either switched back or remain on the server that has last been known to be working.

Health monitoring is fully configurable - from the operation that is sent to the time that Virtual Directory Server should wait before considering an operation as "failed" or not. Operations can be retried several times before a server is designated as down.

Alerts and Notification

Virtual Directory Server integrates easily into a monitoring system, as it is able to generate alerts whenever certain events occur, such as servers going down or coming back up. This can be integrated into an organization's monitoring system. Virtual Directory Server also supports SNMP in client and server mode, making it easy to send traps upon occurrence of definable events.