MS Exchange Clustering

Microsoft Exchange plays a crucial role for a business, and downtime due to an exchange server failure can be costly. One method of reducing downtime is to utilise High Availability Clustering, where in the event of one server failing, another will step into it's place and take over. This failover to the backup node should be transparent to the end users, and cause as little or no disruption or data loss.

Nordicmind can provide LifeKeeper High Availability Clustering solutions for Microsoft's Exchange v5.5, 2000 or 2003 in various configurations (WAN or LAN replication and failover, or a more traditional shared storage environment with failover.)

Protection is available for all standard versions of Exchange - it is not necessary to use Enterprise versions to gain the benefits of high availability.

Using LifeKeeper, it is possible to recover Exchange servers within minutes. Recovery can be to a local server or to a remote server in a separate location over a WAN or LAN.

LifeKeeper supports either shared storage configurations (SAN, SCSI, NAS etc) or use data replication (over a network) for data availability.

Clustered Exchange using Data Replication over a WAN or LAN

The data replication capabilities within SteelEye LifeKeeper for Exchange are based upon the SteelEye Extended Mirroring product.

The advantage of this solution is that it removes the need for an expensive shared storage solution, and also allows nodes to be geographically apart, as long as there is sufficient bandwidth between nodes.

SteelEye's data replication is designed to ensure that an exact copy of the data is present on the target volume, and because it copies at the block level, an accurate and exact copy of the data is maintained, regardless of file system or what files are in use.

SteelEye LifeKeeper for Exchange is an industry-proven solution that replicates data from a local disk located on an active server to a local disk located on a backup server. (Figure 1)

Figure 1: Data Replication Configuration

The recovery process from a local system failure is virtually immediate because data is available on a hot disk.

Data replication can be used to create a range of automated backup procedures that eliminate the need to disrupt users' work activities, or to schedule midnight backup sessions.

Data replication works effectively with tape backup systems to support your organization's established archiving procedures.

Clustered Exchange with Shared Storage

Alternatively, a shared storage solution based around a Storage Area Network (SAN) or Shared SCSI array can be used to make data available to all nodes within the cluster. The advantage of this configuration is that there are no time consuming resynchronisations required if a server unexpectantly crashes, and there is no performance impact associated with synchronous writes taking place.

Shared storage solutions typically cost more, but can provide greater performance and scalability when compared to a single volume being replicated using the data replication solution, however the distance between nodes is normally limited to being within the same room or building, offsite recovery is normally not an option, unless Fibre Channel is being used (which is costly to implement).

Shared Storage with Offsite Replication

LifeKeeper allows for a solution where the data is locally accessible through the shared storage array, and is simultaneously replicated to a remote site. This has the advantage of giving you a local failover node, and a remote failover node.

In such configurations, failover will always go to the local node, which provides the best performance for clients. In the situation where a site failure occurs, the remote node can come online to provide access to other remote clients.

This solution gives you the performance of a local failover, and the security of an exact copy of the data being offsite, and the ability of the remote machine to take over should both local machines fail.

Application Failover

The failover capability within SteelEye LifeKeeper for Exchange monitors Exchange server system resources and services, and initiates a recovery action in the case of a failure. By default it will attempt to recover services locally, however should this fail, a failover to a backup node will take place.

The combination of failover and data replication is the perfect solution to greatly reduce the risk of downtime on lower-end systems that do not use RAID technology. (Figure 2)

Figure 2: Data Replication and Failover

The combination of SteelEye LifeKeeper for Exchange with shared storage is proven clustering solution that eliminates single points of failure in your Exchange system, thereby reducing downtime and enhancing availability (Figure 3)
Figure 3: SteelEye LifeKeeper for Exchange clustering

SteelEye LifeKeeper for Exchange addresses the challenge of planned outages by allowing the manual failover of services to a backup server when maintenance and upgrades are to be performed. This enables user access and operations while planned maintenance is performed.

Disaster Recovery

The Disaster Recovery capabilities within SteelEye LifeKeeper for Exchange provide a complete and unmatched automated disaster recovery environment. (Figure 4).

Figure 4: Disaster Recovery

(Click for full size)

The diagram shows an office environment where one Exchange server is hosted locally, and a disaster recovery server is hosted in a data center.

Such a solution gives enhanced data protection, and service protection after a site failure. Futhermore, it allows for maintenance to take place on the local node during working hours, as clients can continue to connect to the remote site without realising a switchover has taken place.

Our solutions do not require clients to reconfigure, and also allow for failover and failback to occur without intervention (unlike other solutions on the market). These features reduce the cost of ownership and allow for the service to move from node to node as required.

LifeKeeper provides automated and integrated data replication and failover to a remote site:

• Asynchronously and reliably replicate data over WAN
• Reliably monitor critical resources across WAN
• Switch over to remote site only in disastrous events
• Choice of full automation or manually-activated “panic button”
• Access to Exchange Server services is restored in minutes or hours, rather than days

Graphical Interface and Administration

SteelEye LifeKeeper for Exchange simplifies configuration and management of the application failover and data replication processes through the availability of graphical user interfaces for each process set-up.

The first image below shows the Data Replication software in use.

This next image shows LifeKeeper's Java based GUI, which contains a tree like hierarchy of resources with their status shown on each node. Currently Exchange is running (as indicated by the fact all the resources are marked green) on the first node (Jerry). The other node (Tom) is acting as a hot standby.

Services protected

LifeKeeper can be used to protect the following services :

Core Services	Optional Services
Routing Engine	IMAP4
System Attendant	Pop3
Information Store	Event
Simple Mail Transfer Protocol (SMTP)	Message Transfer Agent (MTA Stacks)
WWW Publishing Service	Microsoft Search
	Connectivity Controller
	Connector for Lotus Notes cc:Mail
	Connector for Lotus Notes
	Router for Novell:GroupWise
	Connector for Novell:GroupWise
	Chat

Regardless of configuration, automatic failure detection and recovery is provided. Third party software components can be easily protected (for instance SPAM filters, mobile device controllers, virus scanners etc).

Wish to see a demonstration?

If you'd like to see Exchange and LifeKeeper in action, please contact us to arrange a suitable time for the demonstration.

---

Original Source: www.openminds.co.uk