VMstore™ VM Scale-out
Scale-out storage made simple
Overview:
Scale Out Your Intelligent Storage as Easily as Your Compute
Experience Different! Forget standard infrastructure scale-up and scale-out solutions. They are expensive, hardware-dependent, and require a team of storage PhDs to manage them. What if you could scale-out your storage easily and intelligently, in the same way you scale-out compute in your virtualized environment? With intelligent infrastructure you can simply add another server and let the hypervisor manager optimize VMs across the pool. VMstore VM Scale-out, an intelligent infrastructure technology within Tintri Global Center1 makes this possible.
| Currrent Issues | Recommended Actions | Outcomes |
|---|---|---|
| VMstore 01: High Latency | Migrate 9 VMs from VMstore 04 | VMstore 01 Performance |
| VMstore 02: Running out of Space | Migrate 4 VMs from VMstore 06 | VMstore 02 Space Usage |
| VMstore 03: Running out of Space | VMstore 03 Space Usage | |
| VMstore 04 Load | ||
| VMstore 06 Flash Hit Ratio |
VMstore treats up to 64 systems as a federated pool of storage, and underlying offload technology leverages Tintri replication to move VMs of any size or quantity in minutes. Machine Learning optimizes VM placement, so all you need to do is hit “execute.” Start with one 19TB array and scale up to 40PB with ease.
- VM Scale-Out looks back 30 days to capture performance peaks, not just averages, for better recommendations.
- Machine Learning recommends the optimum placement for each of your apps to minimize impact on capacity and performance.
- VMstore offloaded data transfer for storage live migration reduces lapsed time from hours to minutes. Just hit “execute” and VMstore does the rest.
- Attach persistent policies to apps. Data protection and QoS policies you set for 500 applications remain intact as you grow past 5,000 applications.
1 VM Scale-out requires Tintri Global Center Advanced software.
Experience Simplified, Intelligent Workload Scaling
Standard infrastructure scale-out was not designed for today’s virtualized applications. VMstore intelligent infrastructure was designed to simplify workload scaling—with our flexible architecture and VM Scale-out technology. Here are some VMstore VM Scale-out advantages:
Federated Storage Pool - Treat multiple systems as a pool of storage, simplifying management, planning, and resource allocation. You can manage up to 480,000 VMs from one central console.
Extreme Scalability - Start small, with as little as 17TB and a couple of hundred VMs, and scale to 480,000 VMs or more using the same storage team you have today. The separation of control flow from data flow ensures massive scalability and low latency.
Interoperability - Scale from department to large enterprise with a combination of new and existing systems - all flash or hybrid, partiallyor fully populated.
Guaranteed Performance - Deliver consistent performance for virtualized applications instead of relying on inflexible hardware solutions that can’t guarantee performance and require proprietary interconnects and specialized vendor training.
Grow Performance and Capacity Independently - Optimize your VM placement based on a complete picture of their performance and capacity needs. Scale compute and storage independently to meet specific application requirements. Optimize your resources and avoid overprovisioning.
Experience Differentiated VM Scale-out Analytics
Standard infrastructure analytics tend to be focused on what happened in the past. VMstore is continuously analyzing past and present data to deliver real-time analytics and insight.
Issue Identification - VM Scale-out leverages 30 days of historical data and then projects forward a week, identifying inefficiencies and looking for ways to optimize VM distribution within a pool of VMstore systems.
Recommended Actions - VM Scale-out finds the best location for every individual virtual machine (and the application it supports). It offers you least-cost recommendations to maintain optimal VM distribution and save you time, bandwidth and capacity. VM Scale-out enables you to review and edit recommendations and then view the outcome before committing. Plus, it automatically learns each time its recommendations are edited.
Predictions and Results - If you accept the recommendation(s), VM Scale-out tells you what the result will be and how long it will take to execute. With storage live migration offload, the data moves straight from VMstore to VMstore, maintaining dedupe and compression with zero impact on the hypervisor host, so you can execute the recommended action(s) in a matter of minutes instead of hours. VM Scale-out also offers you the option of selectively not migrating specific VMs.
Documentation:
Download the Tintri VM Scale-out Tech Brief (PDF).