Rencanakan Backup dan Restore Anda
A Planning Guide for Backup and Restore: From Every Day Casualties to Extreme DisastersWritten by Chris Schin, ZettaOctober 3, 2014
High-profile disasters like Hurricane Sandy are the consummate wake-up call for companies to finally get serious about backup and recovery. Yet as organizations look for help preparing for big, unforeseen events, they often overlook recovery practices better suited to the mishaps that are the more common threat to business continuity.
More so than hurricanes or flooding, hardware and software failures along with human error are the primary culprits of system downtime. According to a 2013 disaster recovery report published by a vendor, hardware failures alone constituted more than half of the system breakdowns, while software failures accounted for 18 percent of system downtime, typically due to faulty patch strategies or malware infections. Human error – someone inadvertently wiping out a file system or deleting an important file, for example – was the root cause of 22 percent of downtime compared with those headline-grabbing natural disasters, which were to blame in only 5 percent of cases.
To provide adequate protection, you must be ready for the total spectrum of back-up and recovery situations. Rather than formalizing strategies and aligning with service offerings optimized for disaster recovery purposes, you need to have a back-up and restore strategy that is as deft at remedying the everyday mishaps as it is for handling the least-likely, worst-case disasters.
A Variety of Restore Practices
Regardless of the cause of downtime, there’s no disputing the devastating cost to business. Market research firm International Data Corp. estimates that companies lose an average of $84,000 for every hour of downtime, and it can take up to 30 hours for recovery. These alarming numbers paint a telling picture of what’s at stake for businesses in terms of loss of customers, revenue, and reputation.
Rather than a one-sized-fits-all strategy, there are actually four primary restore use cases that must be handled, each with their own set of challenges and best practices. By understanding the unique requirements of the business and its individual budgetary considerations, you can tailor a back-up and restore plan that mitigates a company’s most significant risks while protecting continuity of operations.
Single file restore. By all accounts, this is the most common situation, where recovery is focused on retrieval of a single email or file that might have been inadvertently lost. Say someone accidently deletes a critical client email or makes changes to a sales presentation and saves it, overwriting the original file. As a result, having an optimal strategy in place should allow for easy and quick retrieval of individual business-critical files.
Best practice: Many back-up and restore offerings store data in a proprietary file format, thus must reconstitute data for retrieval which requires a full restore of an entire data set just to locate one particular file. The primary consideration here is to choose a back-up and restore solution that can replicate data in its native file system format – essentially, a mounted file system in the cloud – like a network file share to locate and retrieve the needed file.
In addition, maintaining an on-site backup of all files is not necessary for this particular use case. Given that the restore procedure involves a small amount of data, cloud backup is more than sufficient.
Corrupted database. Another common recovery scenario is an isolated event where a single database, perhaps one that lies at the heart of a critical business application, gets corrupted or otherwise needs to be restored due to human error and the database is no longer usable. More alarming than the loss of any one file, this situation calls for a recovery option that can facilitate a rapid restore.
Best practice: With a single database, local back-up capabilities would be the recommended approach, especially for organizations that haven’t yet upgraded to fiber-to-premises or Gig-E Internet connections. With a local back-up copy, companies are assured that their systems will be back online with minimal restore time.
Full server. Here’s where things start to get dicey for the business if back-up and restore practices are not properly addressed. The number of servers will vary depending on the size and scope of the business, and certain servers might be home to non-mission-critical systems, thus have more leeway in terms of time to restore. Nevertheless, a proper back-up and restore strategy needs to account for the timely recovery of a full server restore to get a business back on track as quickly as possible.
Best practice: Having an image-based backup of a server ensures that everything, from the operating system to the antivirus software and databases, are fully cloned and ready for operation in the event of a system failure.
This type of backup also involves large data sets so on-site, local backup of the full server image is preferable. A cloud provider that offers a local backup and restore capability – that is WAN-traffic optimized to recover large data sets quickly over the Internet – is also a viable solution.
Office disaster. This is the restore scenario that keeps IT managers up at night and prompts companies to finally look for a back-up and restore partner. In the rare case that a cataclysmic event like a typhoon or fire compromises a company’s operation, business will be shuttered unless proper accommodations have been made to run a backup of core systems and data from an off-site location.
Best practice: This is the one case where on-site local backup, including back-up appliances is of no use, especially if an office or a building is destroyed. In this example, any data or applications that are essential to running the business must be replicated offsite for optimal restoration. Given the scope of this kind of recovery effort, companies need to make sure their restore solution is also WAN-traffic optimized and can support larger data sets.
You need to have a full understanding of the business requirements in order to structure a restore strategy best suited for individual customers. Some companies will want a solution optimized for all or several restore scenarios, while others will have a more targeted set of needs that works best with a particular restore use case.
Restore Use Case
If the cloud provider can provide direct access to data in native file format
Useful for disaster recovery, but not optimal if all you have is a DB corruption
If the cloud provider offers a restore capability that is WAN-traffic optimized to recover large data sets quickly over the Internet then sending server images to the cloud is good
This is particularly optimized if the cloud provider has a Recovery-as-a-Service offering
Best practice for DB recovery is to have a local backup for rapid restore
On-site backup of full server images is preferred when you want to restore the server locally
On-site backup is of no use if the office is destroyed
A Universal Checklist
Let’s take a deeper look at why a one-size-fits-all restore solution isn’t always optimal. An e-commerce company, for example, would be best served by a solution that prioritizes a quick local database restore so they can keep operations running and protect against loss of sales. For a mid-size accounting company where regulatory compliance is an issue, a restore strategy optimized around off-site backup would be the better approach given the need to access back tax records on a periodic basis. At the same time, a law firm might be less concerned about rapid restore of a corrupted database or business system, and instead place a higher priority on instantaneous access to single files in an effort to meet its restore needs.
While best practices will vary according to the restore use case, there are a number of universal requirements to keep in mind when evaluating back-up and restore solutions. Among them:
- Performance. Restoring a file over the Internet is one thing, but recovering a sizeable database or performing a full server restore requires quite a different level of performance. Given that all file transfer and network protocols are not created equal, you should seek a solution that is WAN-optimized, leveraging capabilities like advanced data compression and multi-threaded data transfer to move large amounts of data quickly and securely over the Internet. In fact, Forrester’s top technology trends for 2014-16 reports that the benefits of the cloud will be limited by the speed with which traditional applications are re-written to take advantage of cloud. Without this redesign, benefits will be limited.
- Security. The Internet can be a dangerous place, so it’s critical that an enterprise-ready cloud back-up solution supports the proper encryption and security standards. Look for back-up and restore solutions that support encryption both in flight and at rest and WebDAV (Web-based Distributed Authoring and Versioning) to ensure that large amounts of data can be sent online in a secured fashion.
- Support. Whether it’s online or on-premise, all back-up and restore offerings are going to require some level of handholding. Back-up and restore vendors offer a range of approaches and fee structures for support. Some vendors will charge a premium monthly fee for support while others, especially those with a back-up appliance included with the package, will charge a percentage of the yearly cost (up to 25 percent) for technical support. You need to actor in support costs as part of an overall evaluation.
With global competition raising the stakes across industries, no one can afford to be lax about recovery. Yet equating effective back-up and restore practices solely with disaster recovery is to miss out on a broader opportunity to help companies take on those everyday, man-made mishaps that are the real threat to business continuity.
Chris Schin is responsible for coordinating all product-related initiatives at Zetta. This includes product strategy, direction, and marketing, as well as business model and go-to-market process definition. Prior to joining Zetta, Schin was acting GM and senior director for Symantec Protection Network, Symantec’s Software as a Service platform. Prior to joining Symantec, Schin launched and ran a highly successful consulting practice helping technology startups with market research, strategy creation, TCO and competitive analyses. He has also served as director of product management at both Blue Titan Software and Intersperse Software.