Failover Redundancy and Load Balancing Methods in Exchange Server 2010

Microsoft Exchange Server 2010 provides us lots of tools for failover redundancy (in case of server failure) and load balancing (distributes traffic between servers). In this article, we will try to know a little bit about on these methods to get an idea of them:-

1.       High Availability and Site Resilience

Exchange Server 2010 offers us Database Availability Groups (DAGs) where we can deploy two Active Directory sites in two different locations, and keeping their mailbox data synchronized. In case one site fails, other site will take over its role. We can achieve high availability by configuring up to 16 mailbox servers for automatic recovery without installing any new servers. DAGs provides us local and site-level high availability.

2.       Online Mailbox Move

We can now move the mailboxes whilst the users are logged in. This is only possible if source mailboxes are residing on Exchange Server 2007 SP2 or later, or Exchange Server 2010. It will give users the continuous access to their mailboxes while their mailbox is in the moving process.

3.       Shadow Redundancy

It protects the messages when they are in the deletion process. Before they are actually deleted from the transport database, the transport server first verifies that all the next hops for the message have been completed. If next hop is not completed, the message is resubmitted for delivery to the next pending hop. This is very similar to Transport Dumpster. In shadow redundancy, the source server will delete the shadow copy of the message after it receives the discard status notification from the next hop.

4.       Load Balancing

As the name suggests, it provides balancing of network load or traffic across different servers. Unlike previous Exchange versions, where Mailbox servers were directly connected with the Outlook, Exchange Server 2010 Client Access Server is now used for connecting internal or external Outlook connections.

5.       Block mode

It was introduced in the Exchange Server 2010 SP1- which reduces the exposure of data loss during failover by replicating all logs to the passive database copies in parallel to writing them locally. In other words, it reduces the possibility of any data loss during a failover and the time it takes to perform a switchover.

6.       Multiple Exchange Server roles can now reside on servers providing high availability, thereby reducing servers cost.

7.       High availability can be achieved with only 2 servers in Exchange Server 2010, unlike in Exchange Server 2007, where 4 servers are required.

8.       Exchange Server 2010 added more new options in the Exchange storage menu, like SATA hard disk drive and raid-less- JBOD. Application-level redundancy allows RAID-less (JBOD) storage configuration by replicating 16 copies of each mailbox database resulting in cost savings.

9.       Exchange Server 2010 reduced administrative overhead cost by eliminating the need for creating failover cluster to achieve high availability when Exchange 2010 High availability feature is enabled.

10.   Database Availability Groups (DAGs) provides failover on individual mailbox database level, but not on storage group level which takes time to move the failed mailbox server to the other clustered node or other option was to restore from the backup leaving users offline for few minutes.