Every few years the backup consolidation project shows up on IT’s whiteboard. The objective of this exercise seems noble enough; consolidate point solutions so that the data protection process can be simplified, while at the same time increasing protection quality and lowering costs. But most backup consolidation efforts unravel within a few years. The Achilles heel of backup consolidation is change, and many small to medium sized enterprises (SME) have given up hope of ever creating a single consolidated data protection solution.
The problem is that change works its way into the SME and those changes impact data protection, which often leads to special one-off solutions. Eventually when enough of these special data protection demands are implemented, the backup process is once again fragmented.
SMEs may have one last hope at backup consolidation; the cloud. The cloud has matured to the point that it may be able to meet all of the data protection demands, from basic backup to high availability to disaster recovery and even long-term archival. This article will explore what will be required of cloud-based data protection solutions for them to finally meet the enterprise’s desire for a consolidated data protection strategy.
Agents of Change
The first step in a renewed consolidation effort is to understand what has caused fragmentation in the past. Recent years have brought more change to midmarket enterprises than potentially any other time segment before it. Initiatives like desktop and server virtualization, demand for rapid recovery from any disaster, the rapid growth of data, cloud-based applications, the increasing mobility of the workforce and the proliferation of remote/home offices have driven IT departments to look for new solutions. These change agents include; specialized applications, remote/home offices and workforce mobility.
Most companies had a backup for their “normal” servers, but if the company had applications that required the backup/recovery system to have specialized knowledge about those applications, they often needed a separate solution. For example, it wouldn’t make sense to restore an entire email database just to recover a few lost emails. And especially if the email database is only backed up nightly, restoring an entire database means potentially overwriting all other users mail from that day. Rather, the backup/restore system should know about the email system and be able to navigate the email database backup to recover just the lost mail. Application-aware backup/recovery applications captured application data more frequently and could return the application to operation faster than a recovery from the generic backup process could. But this created another backup system to manage and it meant more hardware and software fees to pay.
One of the results of the internet has been the widespread adoption of work from home initiatives by companies looking to save on office real estate costs, attract top talent by giving them flexible choices on where to live or avoid long commutes to work and more rapidly and flexibly grow and change their field organizations in response to market conditions. This completely changed the network capabilities assumed by traditional backup solutions. A local office LAN, a server room, a local administrator, a persistent connection to all devices to be backed up – all of those factors assumed by traditional backup solutions are no longer valid.
Then there was the problem of protecting laptops. If the organization was protecting laptops at all, it was more than likely using a separate application or service. And, of course, this was prior to the emergence of smartphones and tablets as alternatives to laptops. Now the mobile worker typically has three or four personal devices. Each of these has the potential to create data relevant to the organization that goes unprotected. All of the issues with remote offices apply here as well, plus the fact that these devices are very prone to loss and theft.
For backup consolidation to be successful and not rapidly become obsolete it should match the evolving IT architecture. This means the following requirements:
- It should be extensible for new specialized applications as they arise. This reduces the need for IT to make predictions about the future – which is a risky proposition. Extensibility can be seen by looking at how well and how quickly the current solution is protecting the “new” enterprise right now. Does it provide adequate coverage of recently-installed specialized applications and new virtual infrastructures? How long did it take the solutions under consideration to meet these current demands?
- It should assume that the IT architecture is disparate, decentralized, with areas of high bandwidth and other areas of low bandwidth connectivity, and non-persistent connections. This means that backup jobs should be initializable from either the target-side, as with traditional appliances, or the source-side, as with traditional consumer backup software.
- It should leverage the internet for connectivity just like the source devices do, performing backups and recoveries directly through the internet. But the consolidation solution should also leverage an on-premise appliance when internet latency can be avoided.
- It should add special capabilities to help address risks of data loss with mobile solutions like remote wipe.
- It should be able to recover data rapidly. Meeting today’s protection objectives requires providing solutions that can offer a variety of recovery point and recovery time objectives, from minutes to a few hours. To accomplish this requires intelligently setting applications in a backup mode to access data in a good state so backups can occur frequently without impacting application uptime. It also means minimal data movement, or some form of block data transfer so that data can be copied quickly across the local network and efficiently across the WAN to meet off-site data movement requirements.
Can The Cloud Consolidate Data Protection?
The cloud as an infrastructure is uniquely positioned to meet many of the backup and recovery objectives of today’s enterprise.
The cloud, by its nature, is a decentralized infrastructure, accessible via the internet by any point around the globe. It is accessible by low and high bandwidth connections that may be persistent or intermittent. It is accessible by every form of computing device, from phone to server. Depending on the vendor, the data centers that compose any particular cloud are also decentralized, providing local proximity if required.
Because it is universally accessible, this means that a cloud-based data protection can be managed and monitored from anywhere. This allows disparate locations to be managed from one consolidated IT location, further supporting a consolidation effort.
But “being” cloud is not enough. A cloud data protection solution that offers to consolidate the entire data protection process has to address specialized applications such as email, database and collaboration systems, and protect both physical and virtualized environments. It also should provide special protection functionality for mobile endpoints, whether they are inside the organization’s firewall or traveling around the world. This could include the ability to remote wipe data in the event of loss of an endpoint and geofencing where data can be accessed.
Enterprise data isn’t distributed evenly across the enterprise. There are many locations with low amounts of data where backing up data directly to the cloud is a perfectly viable approach. But there are many locations which will have very large amounts of data, and where backing up data directly to the cloud will take so long due to low internet bandwidth that it will not meet recovery time objectives. Therefore, a cloud data protection solution that is attempting to consolidate the entire data protection process should be hybrid in nature, leveraging a local backup appliance as well as the cloud provider’s facility, as Storage Switzerland discussed in its recent article “What is Hybrid Cloud Backup?”.
A hybrid cloud backup means that the cloud should be able to leverage the presence of a locally installed appliance so that large data sets can be backed up and restored quickly, with the cloud serving as a secondary site in case the appliance is unavailable. The architecture should leverage the infinite storage scalability of the cloud so that local backup data growth does not add to the already daunting data growth problem. Finally a hybrid solution should provide the ability to provide application failover capabilities both at the appliance and in the cloud so that expensive standby servers are no longer needed.
Change is coming though and true consolidation via the cloud may be at hand. Companies like Infrascale are providing data protection solutions that provide hybrid-based cloud data protection that meets various data recovery objectives, as well as protecting endpoints and providing cost effective long-term storage of data.
Leading the backup consolidation effort in an organization used to be similar to plugging holes in a dike. There was always a special exception that the backup consolidated solution could not cover. That lead to fragmentation, and once fragmentation started, the entire effort would crumble under its own weight. The cloud provides the infrastructure that has long been needed to make consolidation a reality. The final ingredients are adding the right software and the right people to make that infrastructure live up to its promise.
Sponsored by Infrascale