Folks from EMC and IBM started Kaminario because they felt passionately that any SSD device worth its name must have an architecture designed from the ground up for fast SSD, not a slow legacy disk architecture tweaked for SSD.

Enterprise mission critical applications need bulletproof reliability as well because any interruption or data loss, even for an hour, can squash not only productivity, but revenue. So Kaminario designed DataProtect’s high availability (HA) and data protection features from the ground up for fast SSD once again.

This is important to remember because many SSD solution vendors claim enterprise-class HA and data protection features, including all kinds of redundant components. But when they start with their HA and data protection checklist you should ask the following questions.

Is HA non-disruptive? Take a look at this Kaminario K2 High Availability Demonstration on the Kaminario Web site. What happens when someone pulls an entire storage node out of the K2? Well, not much and a whole lot. In a matter of seconds, SPEAR’s clustered N+1 architecture detects the outage, starts to fail over to a spare data node, reconfigures the array around the outage, and returns performance to its previous level, all with no data or access loss. Can those other SSD vendors do that? Keep in mind that in a dual-controller configuration typical of other SSD solutions, performance will suffer significantly when one controller drops out and will stay that way until it is replaced. Watch our demonstration video again. Do you want anything less for your revenue critical apps? All of the K2’s storage nodes are also hot swappable, so replacing a node will be non disruptive as well.

Read the rest of this entry »