E-Handbook: How hyper-converged secondary storage fits in the HCI world Article 2 of 4

Nmedia - Fotolia


Craft a converged secondary storage strategy with care

Secondary storage vendors have been offering converged technologies and products for a few years now, and the systems have become more feature-rich and cloud-capable.

These days, you may be hearing about converged secondary storage systems that consolidate data protection capabilities into a unified platform. These products eliminate the need to deploy multiple systems in order to back up or archive data, implement disaster recovery or take other protective steps.

These systems borrow a core concept from hyper-converged infrastructure. They often include virtualization, storage and management software. But while HCI mainly converges infrastructure -- combining storage, compute and virtualization in one box -- converged secondary storage handles multiple use cases such as backup, disaster recovery, archiving, analytics and copy data management.

Although many products share similar qualities, they also differ in significant ways, including the extent they incorporate convergence or hyper-convergence. Here we look at four secondary storage offerings that each take a unique approach to data protection.

Asigra Converged Data Protection Appliances

Asigra sells a line of converged secondary storage appliances developed for service providers that offer cloud backup and recovery services. The appliances are multi-tenant, scalable systems that use commodity hardware such as Intel Xeon processors and Seagate drives, as well as open source software such as FreeBSD and the Z File System, or ZFS.

An established backup vendor, Asigra's appliance focuses more on backup and recovery than other types of secondary storage. The convergence is more about combining of all types of data protection -- for enterprise and cloud-native applications Docker containers and endpoint devices -- into one appliance and management console.

Each appliance is prebuilt and preoptimized for data protection workloads and delivered to the service provider ready to implement. Administrators need only plug in the appliance and run the setup wizard. The appliance provides a centralized web-based management console for viewing and administering the platform components. Once the appliance is set up, the service provider can offer its customers backup and recovery services almost immediately.

An Asigra appliance provides a data repository that can support corporate data from a wide range of sources, including physical and virtual servers, Docker containers, database management systems and on-premises applications. The appliances also support data from software-as-a-service applications, such as Salesforce and Office 365, as well as from PaaS infrastructures, such as Microsoft Azure and AWS.

Asigra currently offers four Converged Data Protection Appliances that differ in terms of processing power, available RAM, caching capabilities and the amount of usable backup capacity. The low-end appliance provides about 10 TB of capacity, and the high-end provides about 360 TB. The appliances use agentless backup mechanisms to securely retrieve, compress and deduplicate the data.

Cohesity DataPlatform

Cohesity DataPlatform is a comprehensive data protection system that provides backup, archival, replication and disaster recovery services. DataPlatform also supports in-place search, real-time analytics and the ability to provision test and development environments. The software fully encrypts, compresses and deduplicates data at a global scale, with support for unlimited snapshots and clones. DataPlatform is based on a secure, multi-tenancy architecture that runs on hyper-converged nodes. The product comes in two editions: Virtual and Cloud.

The DataPlatform Virtual edition consolidates storage and data protection in a virtual appliance based on VMware vSphere. The appliance includes the hardware and software necessary to provide remote and branch offices with backup and recovery services. The converged secondary storage appliance supports geo-replication to centralized data centers as well as native cloud integration with leading cloud providers such as AWS, Azure and Google Cloud, allowing an organization to create a single data fabric that connects edge, data center and cloud environments.

The DataPlatform Cloud edition is deployed on clustered virtual machines in the public cloud, making it possible to create a multi-cloud data fabric that spans an organization's data centers and public cloud services. The Cloud edition supports replication from the private data center to the public cloud, with the ability to control data across all sites. Administrators can use the same DataPlatform UI and APIs to manage both Cloud clusters and Virtual appliances.

Integrated into DataPlatform's core structure is DataProtect, a set of backup and recovery technologies that Cohesity also offers as a separate software product. DataProtect supports policy-based automation, allowing administrators to create application policies that specify requirements such as retention period, archival targets and recovery point objectives. DataPlatform includes a set of REST APIs for managing data protection and integrating DataPlatform with existing orchestration and DevOps tools.

HPE Recovery Manager Central

Hewlett Packard Enterprise (HPE) Recovery Manager Central (RMC) is data protection software that offers snapshot, replication and backup services for application data stored on HPE 3PAR arrays, Hyper Converged 250 systems or StoreVirtual Virtual Storage Appliances (VSAs). RMC acts as a single integration and control point for unifying data protection that can span all three platforms.

Although many converged secondary storage products share similar qualities, they also differ in significant ways, including the extent they incorporate convergence or hyper-convergence.

RMC makes it possible to back up the data directly from the supported platforms to the target storage, without affecting the application servers. The software can also back up data to on-premises object storage, as well as to public, private or hybrid cloud platforms. To support these processes, RMC uses technologies such as 3PAR SnapDiff, which carries out incremental backups by detecting changed blocks.

In addition to deduplicating the data during backup, RMC uses an array-based approach to creating space-efficient, self-contained snapshots that can be recovered to any point in time. Because the snapshots are independent volumes, they can be restored to the original 3PAR array or to a different one. RMC also provides the Peer Copy feature for copying data bidirectionally between 3PAR arrays and StoreVirtual VSAs.

HPE also offers a set of RMC application plug-ins for integrating with systems such as VMware vSphere, Microsoft SQL Server, Microsoft Exchange, Oracle Database and SAP HANA. In addition, HPE provides the APIs necessary for developers to integrate RMC's data protection capabilities into their own applications.

Rubrik Cloud Data Management

Rubrik Cloud Data Management (CDM) provides backup, recovery, archival and disaster recovery services, along with compliance, analytics and search capabilities. CDM also supports testing and development scenarios. Organizations can deploy CDM on plug-and-play appliances, Rubrik's Edge software appliances or other certified hardware platforms. In addition, CDM can run on cloud platforms such as Azure, AWS or Google Cloud.

At the heart of CDM is a software-based architecture that decouples application data from its infrastructure to create a single software fabric. Rubik has also built intelligence into the software to improve storage and data management efficiency.

Rubrik created CDM as a purpose-built system for cloud integration and unlimited scaling. CDM can capture data from both physical and virtual environments hosted on-premises or on cloud platforms, whether private, hybrid or public.

CDM includes a number of features for unifying and managing the disparate application silos. For example, CDM implements a flash-optimized distributed file system that is fault tolerant and linearly scalable. CDM also employs a distributed metadata system that works in conjunction with the file system to provide a global index and catalog.

Another important component is a data management and global search feature that offers data lifecycle management. CDM also includes a distributed task framework engine that assigns and executes tasks globally across the cluster, as well as a policy engine that orchestrates service-level agreements across the data lifecycle. CDM provides a number of other features as well, making it a top contender in the converged secondary storage market, along with Cohesity DataPlatform.

Dig Deeper on Converged infrastructure management

Cloud Computing
and ESG