So, you’ve decided noisy neighbors are threatening the performance of your mission-critical apps that are running on your Cloud Service Provider’s (CSP) multi-tenant infrastructure. And of course, you want to get onto your own hardware via your CSP’s Private Cloud offering.
Unless your CSP’s Private Cloud is built on Tintri storage, you may be spending a lot of time and money for no net gain in performance. That is to say, you may be your own noisy neighbor if your VMs are running on traditional storage.
Consider that Tintri is the only storage platform that is purpose-built from the ground up to be VM-aware. That means that Tintri’s unit of visibility and control is the VM—not the LUN, not the volume. Tintri analyzes and dynamically adapts to the performance of every VM on the system and—this is the important part—automatically provides workload isolation at the virtual disk level.
Compare the following scenarios:
Scenario 1: If you place two of your busiest database servers on a traditional array, they can both create contention and performance degradation for each other, and the VMs that also live on that array. (Precisely the thing you had hoped to alleviate by moving to “dedicated” hardware.)
Scenario 2: If you place those same workloads on Tintri storage, then the storage will automatically, and adaptively provide isolation for those VMs and the VMs reside there with them, such that their performance demands are met without impeding each other.
As you can see, there is little benefit to moving your VMs into a dedicated environment (Scenario 1) when VMs can still affect each other on traditional storage.
Ask your Service Provider today if they offer Tintri-based Private Cloud. (And while you’re at it, ask them if they use Tintri for their multi-tenant Cloud offering—per-VM isolation of noisy neighbors works there, too!)
Interested in learning more? Check out our dedicated Private Cloud page for more information. And while you're at it, explore some of our other resources below.
Unique control with VM-level actions for infrastructure functions including snapshots, replication and QoS make protection and performance certain in production, and accelerate test and development cycles.