The process of copying data into backup media can be traced back to the early days of magnetic tape technology, around the early 1960s. Through the years as advances in technologies has created faster and more reliable computing systems, backing up data has remained a constant practice for organizations regardless of size and industry.
Data&StorageAsean recently had the opportunity to speak to Charles Clarke (CC), technical director for Asia Pacific at Veeam Software to discuss the rapid changes within the IT industry and how this is impacting backup and recovery practices within the enterprise. In particular, how virtualization technologies is impacting data protection, including backup and recovery.
DSA: How important is backup (as a process) in the data protection strategy of an organization?
CC: Backup is only tactical but recovery is both strategic and critical. Every backup process should have a clear and regularly tested recovery plan that is well understood at all levels in the organisation. According to Gartner two out of five SMB organisations that have a major outage never recover and every organisation will have at least one outage of some level over the course of a year. Having a good recovery plan built on the right tools might mean survival.
As VMware’s dominance in the server virtualization space erodes with the success of Microsoft and Red Hat, it is benefiting users by giving them choice and flexibility from a commercial standpoint. Clarke concedes that from a feature standpoint there's not much to choose from at the Hypervisor level so decisions will hinged upon management tools, cloud integration and other things that may future proof the IT direction of the organisation.
DSA: How does this trend of multi-visor implementation impact the cloud?
CC: The biggest impact is to cloud service providers themselves. Remember that the majority of cloud service today is delivered via managed service providers or datacentre service providers. Having the ability to use a Hypervisor to match their business needs makes them more economical and flexible. It should be transparent to the end user, but the end user is still responsible for that data so should hold the cloud provider accountable for backup and DR. It also allows them to align themselves to particular customer scenarios e.g. VMware to VMware replication using Veeam, or recovering hyper-v workloads to a hosted hyper-v cluster.
DSA: How does this become a problem as it relates to backup in the cloud?
CC: Recovery is ultimately the responsibility of the CIO. They need to hold the cloud service provider accountable and ensure tested, clear recovery plans. With the right tools the Hypervisor can be commoditized.
DSA: If a company decides to migrate parts of its storage needs to a hosted provider, keeping mission-critical data on-premise, what is the right backup strategy for them?
DSA: The key element of this strategy is to use the right tools for the job rather than attempting a 'one size fits all approach'. If the mission critical data resides on a virtual platform then using tools built for virtualisation is the right choice. Using tools relevant for that cloud platform is also the right choice even though it may mean a different toolset. We see organizations using their on-premise infrastructure for DR from the cloud too. The cloud service provider should prove recoverability regularly.
DSA: The realities of cloud storage provision today are that cloud storage is still finite; access is constrained by bandwidth; protecting data in cloud is not 100 percent certain; and backing up data to on-premise is still relevant. What are the options?
CC: We recommend that enterprises pursue a multi-tiered approach to data protection. We preach what we call the 3-2-1-0 rule: Maintain THREE copies of data including production; have data reside in at least TWO media potentially it could be tape or cloud storage; keep ONE of those media be offsite; and aim for ZERO error in recovery. The most important aspect of a backup and recovery strategy is actually the recovery, and recovery takes on even more importance particularly when you are considering cloud storage for backup.
There is a lot of interest on cloud storage these days but the reality is that recovering data from the cloud is more expensive and slower than putting the data there in the first place. In our experience as a backup vendor, organizations use cloud storage as an archive tier, and for backup data, they use on-premise to ensure rapid recovery.
DSA: Which types of customers and environments is the Veeam Backup and Replication aimed at? What is Veeam’s licensing model and in what situations is this economical?
CC: We cut across all sectors and verticals, including SMBs. Our target market is everyone using virtualization. Our perpetual licensing model is per CPU. A few years ago we introduced a clone licensing model designed specifically for managed service providers, hosters and datacentre providers – people who are looking to host those workloads.
DSA: Does Veeam support site-level replication and DR for physical and virtual servers in the environment?
CC: Our technology is built only for virtualization. The conversation we have with customers is to offer best of breed for virtualized environment.
DSA: A full data protection strategy includes encryption for data whether it is in transit or at rest. Does Veeam offer this functionality?
CC: Yes, in our cloud edition of our backup and recovery software we include encryption for data both in motion and at rest. However, in our standard edition, our experience is that customers already have a solution in place for data encryption.
DSA: How will Big Data impact the data protection strategies of users and vendors alike? Does Veeam have credible solutions available today?
CC: There are challenges around the definition of Big Data. In my view Big Data is data that cannot be managed in the conventional approach – too big for regular file systems or database. Today a lot of the data that goes into Big Data is not necessarily important data. What we tell customers is that you need to identify what parts of your Big Data are relevant that warrants data protection in the form of backup and recovery. It’s up to IT managers and CIOs to define what needs to be protected, where it needs to be stored, and how much it will cost, as part of an effective protection strategy.
IDC predicts storage will grow by 50 times within the next 10 years. Research has shown that IT staffing is projected to grow by only 0.5 times over that same timeframe. IT departments need to learn how to do more with less, and companies who are using multiple backup and recovery solutions to protect data across their physical, virtual and cloud infrastructures will turn to all-in-one data protection solutions that protect information across heterogeneous environments. As a result, the Purpose Built Backup Appliance (PBBA), an all-in-one backup and recovery solution that customers can plug in and trust to protect rapidly expanding volumes of data across heterogeneous environments, will continue to be in high demand.
In addition to changes around what solutions companies buy, there will also be an evolution in how they purchase. Companies want to purchase technology the way they purchase consumer goods on the Internet. They want to conduct their own online research, and when they’re ready to make a purchase, they want simple pricing and a purchasing process that results in quick time to value. They do not want to be bombarded with outbound campaigns, cold calls and sales pitches.
In 2014, data protection vendors will adapt their sales models to accommodate the way customers want to purchase. Vendors that focus on inbound requests, holding meaningful conversations with prospects and making sure companies are buying through a channel in which they feel comfortable, will have the most success.