Rajesh Ramchandani, GM, CloudPlatform and Services, Accelerite and
Ashok Rajagopalan, Head of Products, Datera
Enterprises are looking to leverage the blueprints of web-scale computing to transform their data centers. Web-scale computing allows data centers to be built with the intelligence for scalability, and reliability in software using commodity hardware. This helps transform these data centers to become more agile. Cloud service providers such as Amazon, Google, and Facebook, are very successfully demonstrating this new paradigm to deliver IT services with high level of availability at a very low cost. Enterprises continue on their journey towards refactoring their on-premise infrastructure based on the blueprints of the public clouds such as AWS, Azure, and GCP.
This journey entails enterprises transitioning to next generation software defined technologies while leveraging investment in existing hardware. Several cultural shifts also are required to achieve this goal towards agile infrastructure. Broadly speaking, some of the core tenants are:
- Software-defined infrastructure provides the ability for customers to leverage industry standard x86 servers with policy-based control software. Customers want to take advantage of the commoditization of hardware, the fast evolution of media, and improvements in speeds and capacities.
- Automation to simplify operations and manage resources as programmable entities. Modern thinking is about holistically managing infrastructure through programmable infrastructure rather than performing operations on a box by box.
- Shared infrastructure as customers want to eliminate existing application silos, maximize utilization and run a wide variety of application workloads with determined service levels.
Accelerite CloudPlatform and Apache CloudStack are mature and proven enterprise cloud orchestration platforms that can operate at very large scale. Cloudplatform is designed to deploy and manage large number of virtual machines, as a highly available, highly scalable infrastructure as a service (IaaS) cloud computing platform. CloudPlatform and CloudStack are used in production by a number of service providers to offer public cloud services, and by many enterprises to deliver on-premises private cloud offerings.
To complement the flexible architecture, Datera is partnering to address the persistence and block storage needs. Accelerite is reselling Datera Elastic Data Fabric (EDF) storage system through their channels as part of the broader joint engagement with Datera. CloudPlatform and Datera as an integrated solution provides lowest total cost of ownership (TCO) for enterprise clouds for any equivalent hyper-scale or hyper-converged infrastructure.
Datera EDF offers Primary Storage which is configured at the Cluster or Zone level and is intended to be root disk volumes and/or data disk volumes for all the virtual machines. Datera’s seamless integration with CloudPlatform and CloudStack makes primary storage management and provisioning simple. For the Datera’s CloudPlatform and CloudStack plugins, please refer to https://github.com/Datera/cloudstack-driver
For ease of deployment and operations, we have a Deployment Guide documenting the various steps to configure, provision and manage Datera EDF with CloudPlatform.
We also have a few short videos to help to operationalize Datera EDF in a CloudPlatform environment here
Volume Service Levels
- Ability to create multiple service levels customized by capacity, IOPS, bandwidth, latency and IO durability
- Ability to differentiate the acquisition cost of the service tiers by placing volumes/replicas on hybrid flash and all-flash nodes using all-flash replicas, primary replica on flash, and all hybrid replicas
- Seamless integration with CloudPlatform and CloudStack environments through the Plugin driver.
- Support fast snapshots through snapshot offloading to the Datera EDF
- Creating volumes, templates and VM instances from snapshots
- Attaching and detaching volumes to VMs on different clusters
- Self-adaptive system which best fit places volumes on appropriate nodes, and actively rebalances the system based on changes in workload and network behavior
- Cloud-based analytics framework to actively monitor the on-prem system, and provide analytics on application and system behavior
We envision a world towards simple, efficient and autonomic infrastructure. We question the assumptions, challenge the operational model and reject the cost structure of today’s brittle, rigid and unscalable IT.