At PrimaryIO, one of the many things we aspire to do is to help our customers realize the full value of their existing investment in storage systems. With technology often becoming obsolete even before it is introduced, and vendors touting All-Flash arrays as the the panacea for data storage performance problems, resisting the urge to upgrade prematurely and unnecessarily is a huge challenge for system admins. This urge is even more prominent in data centers facing pressure to offer high quality services in the presence of increasing load due to virtualization.
We recently had a meeting with a leading financial institution in England. They have Netapp file handlers that are getting swamped IO-wise from the increased density that virtualization has created. They not only need to deliver more IOPS but they also need to realize the full value of existing investments in storage technology. When an organization invests a few $100K in a storage array, they usually anticipate using it for at least 5 years. At this point they are not willing to upgrade to All-Flash arrays prematurely. Furthermore, adding more spindles to increase IOPS also feels like adding more of a bad thing, since they don’t need the excess capacity, and they plan to migrate the storage array once it’s full value has been realized. When it’s too soon for an All-Flash upgrade, and when adding more spindles is also unwarranted, an organization needs to explore the alternative solution of deferring the upgrade with a better alternative.
That’s exactly what our product, PrimaryIO APA, does. It’s a software defined solution that can handle the increased IOP demand by putting the active data set IO on an in-server cache without additional costs to the existing storage platform. Using PrimaryIO APA will allow this organization to realize the full value of their investment in Netapp before migrating the storage system to the latest generation array.
Although the middle path of deferring an upgrade sounds impressive, we are often questioned, What happens after we upgrade? Will PrimaryIO APA be useful even after the upgrade? Our answer is always affirmative, assuring them that we will continue helping in more ways after the upgrade.
There are several ways in which our solution will continue to be relevant, but let’s be specific for now. This organization occasionally gets peak workloads that cause severe performance challenges. Having a PrimaryIO cluster (with in-server SSD cache) readily available will allow a vMotion of this workload to this cluster, resulting in a performance boost and the ability to protect other workloads on the latest generation array.
Since our solution identifies and caches the active data set IO, we will always be able to cache the most frequently used data on a storage medium that is significantly faster than the storage array; boosting even the most expensive storage solution.
Does your data center have a storage performance problem? Are you unwilling to migrate your storage system to a newer technology (yet) and are also hesitating to add spindles to a storage system that has already become old-tech? We will be glad to work with you, understand your storage system, and help you determine the correct path to boost it’s performance. Please don’t hesitate to write to us at firstname.lastname@example.org.