Veeam Backup and Replication brings many advanced data protection and recovery features to virtualized environments. And backup appliances have become a popular way to efficiently store, recover and electronically vault virtualized data for disaster recovery (DR) purposes. The challenge is that most backup appliances don’t provide the integration required to fully exploit all of Veeam’s capabilities. ExaGrid’s backup appliance architecture consisting of a front-end landing zone with scale-out capabilities, combined with the integration of a Veeam backup data mover, enables businesses to fully optimize virtual machine backups while gaining all the storage efficiencies of a scale-out deduplication backup platform.
Virtualized Backup Pioneering
Founded in 2006, Veeam was one of the first software companies to provide a purpose built backup application for protecting VMware virtualized infrastructure. Since their inception, Veeam has enjoyed explosive growth and is recognized as one of the market leaders in the virtualization backup space – for VMware and Hyper-V environments. Likewise, ExaGrid was one of the first disk based backup appliance manufacturers, having entered the market approximately a decade ago. ExaGrid has developed continuous enhancements to their architecture to meet the evolving data protection needs of today’s virtualized data center environment.
Veeam Ready Integration
To deliver increased functionality in Veeam backup environments, ExaGrid EX2000 appliances and above are now integrated with a Veeam accelerator data mover. By hosting the Veeam data mover, ExaGrid appliances provide a direct connection with the Veeam management server that is more optimized than a comparable CIFS or NFS mount point. This connection uses a communications protocol designed for large backup data which improves backup and restores speeds. In addition, for certain functions it also reduces many network hops (primary storage – Veeam backup server – backup network – disk appliance). In addition, this data mover enables Veeam synthetic full backups to be fully processed on the appliance. In other words, all the Veeam incremental backup jobs can be concatenated on the appliance to provide a full backup image. This eliminates the need to run a lengthy secondary process to produce a synthetic full backup image.
Veeam DR Automation
Another benefit of the integration with the Veeam backup data mover is that the Veeam management server maintains an “awareness” of any replicated jobs that have been processed by the ExaGrid appliance. Without this data mover, administrators would have to manually update the local and remote Veeam catalogues each time data was electronically vaulted off site for DR protection. The ExaGrid Veeam data mover, however, automates the replica/catalogue update process and as a result, makes the environment DR ready as soon as the replication process completes. Moreover, by offloading the Veeam backup server from processing replication jobs, there is that much less overhead placed on production VMs.
While this direct integration with Veeam helps enhance the interaction between the Veeam backup server and ExaGrid’s backup appliance, the back-end disk architecture is still a critically important factor for enhancing backup and recovery speeds in Veeam, as well as other backup application environments. In addition, ExaGrid’s backup appliance design provides the right tandem of performance storage and efficient deduplicated storage capacity, to enable advanced recovery features, like Veeam’s “recovery-in-place” technology, while enabling data centers to maintain multiple backup images of their business data on cost-effective storage resources.
Grid Powered Backup Efficiency
ExaGrid’s architecture can best be described as a multi-node, scale-out backup grid that features a two-tier storage design. This architecture combines a front-end disk cache landing zone to support rapid backups and recoveries, along with a second tier of deduplicated storage capacity for long term data retention. This design is a differentiator for ExaGrid in that it enables data centers to get the best of both worlds – a fast disk “landing zone” that is unencumbered by the overhead of data deduplication, to support rapid backup and recovery, and a deduplication tier that can efficiently store weeks or months worth of backup images.
Enabling VM Recovery-In-Place
This landing zone is also proving to be a valuable asset for the recovery-in-place feature that is available in Veeam backup environments. As the name implies, recovery-in-place is a technology that allows virtual administrators to recover failed VMs directly off the backup volume. This can help speed up recovery efforts as administrators would not have to wait for data restore processes to finish before fully recovering a failed application. Instead, it can boot directly from the backup volume.
But as we have discussed in past articles, utilizing recovery-in-place can be a non-starter if the backup appliance doesn’t have robust disk resources to support virtualized production applications. This is where ExaGrid’s front-end landing zone can be a significant differentiator over comparable backup appliances. The landing zone in effect serves as the backup and recovery workhorse while the second tier of what is called “adaptive deduplication”, can ingest and deduplicate data in parallel as backups are pushed to the disk landing zone.
Scale-Out Backup Architecture
The increasing need to scale backup environments to meet the twin forces of highly virtualized environments and unrelenting data growth, makes ExaGrid appliances an ideal solution for the enterprise. The key to maintaining investment protection of the backup infrastructure is ExaGrid’s ability to start small and scale-out into a large multi-node backup storage grid. Each ExaGrid appliance has processor, memory, bandwidth and disk. As data grows, all four resources are added, which provides for a fixed length backup window.
Traditional “scale-up” backup systems support a fixed amount of storage capacity. Since only disk capacity is added to a fixed set of processor, memory and bandwidth resources, as the data grows, the backup window also increases, which can result in missed backup windows. Once the processing power of these systems are exhausted, end-users have to either add a second system into the data center, which has to be separately managed, or fork-lift upgrade into a larger scale-up platform. This increases the total cost of ownership (TCO) of the scale-up solution. In addition, scale-up systems require an up-front investment in all the processing and controller resources of the box. This means that the user will not get the full value of the system until they are utilizing at least 2/3 of the storage capacity of the unit. And by that time, they will be close to needing an upgrade.
ExaGrid’s scale-out systems, on the other hand, can start out with a single node and incrementally scale-out as the backup capacity and performance needs of the environment increase over time. This allows businesses to only invest in what they need today. Furthermore, from an investment protection standpoint, ExaGrid supports the use of multi-generation nodes within the same grid. For example, a generation 2 node will interoperate with a generation 4 node. In fact, ExaGrid still supports their first generation node systems in the field today. And some of these are integrated in the same grid with the latest node architectures.
Scale-Out Investment Protection
In addition to providing better investment protection, the ability to add nodes “on-the-fly” enables data center planners to perform non-disruptive upgrades to the backup environment. This means there is no need to plan for data migrations, which can be a big win for already overworked IT professionals.
Another benefit is that by adding backup resources “just-in-time”, the backup window can, in effect, remain static. In other words, resources can be added BEFORE the backup window grows longer. This means that there is less of a risk of missing the backup window and letting critical business data go unprotected. This also enables businesses to take advantage of ongoing decreases in the price of storage. By waiting to deploy storage purchases until they are absolutely needed, IT organizations can reduce their backup total cost of ownership (TCO).
Enterprise Class Scalability
The highly scalable nature of ExaGrid’s backup architecture enables it to support small, medium and enterprise data center environments. For example, a deployment can start out as small as a single node which can support a 1TB backup environment and can grow into a grid that can support up to 294TB’s of primary storage backup data. Through data deduplication, backup environments can conservatively see 10TB’s of “logical” storage capacity on the entry-level node and up to 2PB’s of logical capacity in a fully configured grid.
ExaGrid models vary in size from 2U, 3U and 4U appliances which contain storage capacities ranging from 5TB to 48TB’s of raw storage. And as discussed earlier in the article, each appliance model can be inter-mixed, with up to 14 in a single Scale-out GRID, regardless of storage densities, giving data center planners the flexibility to increase their backup storage grid as granularly, or as large as they wish, without worrying about node compatibility.
Veeam provides many advanced data protection features for virtualized environments but these capabilities can’t be fully leveraged unless the underlying disk backup architecture has both the native software intelligence and storage design to enable these features. ExaGrid’s scale-out backup architecture, two-tiered storage design and it’s direct integration with Veeam via the data mover, enable Veeam data center environments to cost effectively protect their critical business data now and well into the future.
Watch the webinar on demand "The five ways your Backup Design can Impact Virtualized Data Protection"
Sponsored by ExaGrid