The Power of Autonomous Operation
With autonomous operation, software takes care of infrastructure management tasks. Autonomous capabilities have been available on the compute side for quite a while with server virtualization and dynamic resource scheduling (DRS). Now, containers and advanced schedulers are taking the autonomous operation of compute even further. Network virtualization and software-defined networking (SDN) are improving autonomous operation in the network layer.
Storage, however, is complicated. Among the many variables that storage performance depends on are block size, read/write ratio, sequential versus random I/O, drive type, and controller design. Storage behavior can change dramatically in response to even a modest change in one variable. This is why intelligent storage needs to protect administrators from this complexity, but this also why autonomous operation for storage is not a trivial undertaking.
Autonomous Operation and Tintri Intelligent Infrastructure
Tintri Intelligent Infrastructure is pioneering autonomous operation for storage, delivering capabilities that help your team get more from storage while investing far less time on management. Just like self-driving cars and airplanes with autopilot, autonomous operation frees up your team to do other things.
A key tenet that underlies autonomous operation is predictable performance. Tintri VMstore has designed predictability into both hardware and software. VMstore hardware delivers predictable results despite component failure or during maintenance activities. VMstore software uses a VM-centric filesystem to deliver predictable performance and autonomous operation that includes:
- Auto-tuned quality of service (QoS) automatically keeps “noisy neighbors” from interfering with other workloads so administrators don’t have to worry about load balancing or the placement of different types, sizes, or numbers of VMs inside LUNs or volumes.
- Intelligent flash allocation ensures that 99-100% of I/O is delivered from flash for hybrid storage configurations using built-in analytics to determine the working set.
- VM scale-out does for storage what server virtualization with Predictive DRS does for compute. Once workloads are deployed, the storage learns from the environment. In-depth analytics are used to recommend optimum placement of VMs on storage based on capacity, performance, and flash requirements.
Just as a a pilot can take a plane off autopilot when he or she needs control, Tintri VMstore autonomous operation gives you fine-grained control over the platform when you need it. You’ll hear more about that in upcoming posts.
A Better Approach to Storage and Data Management
At Tintri, our goal is to deliver the simplest management experience we can, eliminating complexity and allowing our storage to operate autonomously and make smart choices for you whenever possible. We think the combination of Tintri VMStore’s autonomous operation and real-time and predictive analytics can transform your approach to IT operations.
Stay tuned for additional posts in this series to learn more about the capabilities of Tintri VMstore QoS and VM scale-out.
The only thing better than simplified IT management is infrastructure that manages itself. Welcome to the era of self-driving storage.
In today’s digital economy, the priority of IT teams is shifting away from infrastructure management to a more intense focus on delivering new and improved applications and services. At Tintri, we believe the best way for enterprises to achieve this is through an enterprise cloud that delivers public cloud agility inside your data center, so you can exercise full control over critical applications, ensure compliance, and better predict and control costs.
Tintri’s Intelligent Infrastructure delivers automation at the right granularity, self-service that allows developers and line-of-business managers to move faster, analytics that provide deep insight into IT operations for faster and better planning decisions, and autonomous operation that decreases management tasks.
When it comes to infrastructure, the only thing better than simple management is eliminating the need for management altogether.