As of today RecoverPoint 3.3 is now GA! So what does that mean for you current and future customers? Well lets see, nothing but an incredible amount of flexibility and control I would say! So what's the goods? This major release touts such new features as Distributed Consistency Groups, Clariion Splitter Sharing, and Automated Failback with VMware’s Site Recovery Manager. Each of these bundles of joy takes you one step closer to replication shangri-la ..this way please –> –> –>

Distributed Consistency Groups

A major component of RecoverPoint has long been the notion of Consistency Groups. This logical entity maintains write order precision across SAN volumes of user defined likeness. This parallelism (so to speak) is important during replication as recovery of an application is usually dependent on all corresponding data LUNs being in lock step with each other.

If one LUN and its data is 2 steps forward and another LUN and its data is 4 steps back..well lets just say recovery just got a whole lot more poo-laden. Now traditionally CG’s have always been defined to a single RecoverPoint Appliance or RPA. And as you can imagine the RPA’s can only push what they can push which is typically around 75MBps sustained, burstable to 100MBps. Which mind you, is nothing to laugh at for a single appliance. If your write rates pushing the boundaries of these limits, you add another RPA… horizontal sizing and scaling is key.

Ok enough mister of this foundational knowledge, what does this really give us. Simple..Your writes are no longer bound to a single RPA you now can distribute those writes for up to 8 DCG’s across up to 4 RPA’s simultaneously. Load is now balanced, throughput has now increased. Consider this graphic, what's circled in red is how CG’s were transmitted in previous versions, source RPA to peer RPA..1 to 1.

Yes, functions such as Group Sets did allow you to maintain fidelity across multiple CG’s, across multiple RPA’s but at the loss of such goodness as snapshot consolidation. With Distributed CG’s you can now move beyond the defined throughput of a single RPA and scale up to potentially 3 times that. That’s a winner in my book ; )

Clariion Splitter Sharing

Anybody who is anybody can appreciate the simplicity in the Clariion Splitter. Elegant, poetic, yet..yet fully functional. 99% of our customers having Clariions, which corresponds to 99% of our RP installs being Clariion splitter based, which corresponds to 99% of our customers having RecoverPoint/SE. And what is a limitation of  RP/SE? Two RPA’s per site and yes..one Clariion Splitter per site. Many-to-One is now available and allows the sharing of a single Clariion among multiple RP (or RP/SE) clusters. Now could this model fit the service provider model, where potentially the provider serves up a Clariion and multiple destination RPA’s and the customer provides the other half of the solution at their production site. That remains to be seen, but none the less a welcome addition.

Failback by way of VMware Site Recovery Manager

This has been on a lot of people’s wish list for a while, to offer this first with Celerra Replicator made sense from a numbers perspective but that's it. What you get is complete automated failback, you failover over, EMC fails you back : ) How much simpler could it be, never a real limitation of SRM always a limitation of the replication engine or perhaps the smarts within the SRA. Enough said…

For a complete list of new features go here.