3 Ingredients to Keep Storage Scale-Out Simple
By Jack Bornhoft
Is your virtual data footprint growing rapidly—or are you expecting to have a growth spurt in the near future? Then it’s time to think about scale. Now, scaling compute is relatively friction-free, but scaling storage on the other hand can be a challenge. That begs the question:
What if you could scale storage as easily as you scale compute?
If you need more compute resources for your virtualized workloads, you simply install a new virtualized server, add it to the resource pool and automatic live migration optimizes the VMs across the pool.
Tintri VM-aware storage has proved to be as simple to scale-out as compute. One Tintri customer can provision more than 50,000 VMs every month and manage 3 PB of data—all with one Windows Admin spending just four hours per week on storage.
What does it take to keep storage scale-out this simple? These three ingredients:
1. Architecture. Conventional storage scale-out requires a significant upfront investment of time and money. You generally have to purchase expensive hardware, proprietary interconnects and even additional software—plus you need to staff experts that can make it all work together.
With Tintri you can start with a single 17 TB VMstore, expand up to 308 TB and scale-out to 10 PB. The architecture is federated—you can manage all your devices (including a mix of all-flash and hybrid storage) as a single pool. And it is loosely coupled—control flow and data flow are separated, ensuring low latency across a large number of nodes. Importantly, all VMstores use the same Tintri operating system, which means your investment is future proof—future releases are compatible with your growing pool.
The bottom-line is that you can start small and expand as needed.
2. Automation. As your footprint grows you need to lean more on automation to avoid scaling your management burden. This is another area where Tintri’s VM-aware storage stands apart from conventional storage.
When you add another Tintri VMstore to your pool, Tintri’s VM Scale-out technology automatically recommends the optimal placement of every VM across the pool. The recommendation algorithm factors in the behavior of each individual VM—required resources, potential for space savings, and the cost in time and data to move VMs. This recommendation engine is always at work making least-cost decisions that ensure the organization gets the most value from its storage capacity and performance.
Another example of automation at work is policy management. Applying policies to a growing footprint is not a trivial task. With conventional storage it’s highly manual, and as you move virtual machines between LUNs or arrays you typically have to manually re-apply policies. Intelligent scale-out helps you organize your virtual machines into similar groups where you can apply quality of service and/or data protection policies. As individual virtual machines are moved within the scale-out environment, they retain their policies automatically. The policies you set when you have 500 virtual machines stay in place as you grow to 5,000. Here’s how it works.
3. Analytics. All storage platforms offer generic analytics of one form or another. The question for the growing organization is how can you put that data to work to make better scale-out decisions.
VM-aware storage platforms like Tintri can analyze the performance of individual VMs to provide a depth of insight not possible with LUN / volume-based storage. For example, while all storage can estimate when you will run out of capacity, Tintri goes a step further by forecasting precisely when you will need more performance and/or working set.
With Tintri Analytics you can define application profiles—this involves grouping together types of applications (for example, all virtual desktops or SQL servers), so you can look at the behaviors of this group. You can see how many resources the application type is consuming, and drill into the member virtual machines to see which outliers need attention. If you’re asked to add another 10 SQL Servers, a quick what-if analysis of the application profile will show you whether you can accommodate the request or if you need more resources.
If your virtual footprint is growing, the above three elements should be sought-after capabilities from your storage to prove that scaling storage can indeed be as simple as scaling compute.
Scott Buchanan leads Tintri’s Corporate Marketing function. His team is responsible for Tintri’s brand, content and social media.