When disaster strikes in the data center, the IT department needs to be ready. Improvising – though it may be necessary – is rarely a good idea when the survival of the business is on the line.
Given the importance, here are eight tips to help you out. (For additional tips, see our earlier article on IT disaster recovery).
Automate Disaster Recovery
When handling day-to-day emergencies, it is easy to forget about the possible disaster that may or may not occur at some point down the road. Then, when a real disaster hits, personnel may be missing, will definitely be under pressure and prone to making mistakes.
“Automated DR understands the precise order and procedures and applies this understanding to the recovery process,” says Ralph Wynn, technical director of product marketing at FalconStor. “It eliminates human error and returns the organization back to normal operations within minutes, instead of hours or days.”
Taking Care of the Physical/Virtual Interaction
With multilayered technology, different parts of a service can be running on VMware arrays, physical appliances, appliances, storage networks and in the cloud. Each of these can have its own backup and restore procedures, all of which need to work together seamlessly.
“Most companies fail to properly protect Physical and Virtual environments correctly, or use a variety of systems that do not communicate together and make it ultimately difficult to recover properly from a disaster,” says Seth Goodling, Director of Strategic Technology for Acronis Backup and Recovery 11.5.
Virtualize Storage and Servers
Server and storage virtualization are great ways to improve the resiliency of systems inside the data center and can be easier to restore at a disaster site.
“Having virtual machines can enable portability to make it easier for relocating either proactively, or via a restoration at a DR site, by abstracting hardware dependencies,” says Greg Schulz, founder of the Server and Storage IO Group and author of the book Cloud and Virtual Data Storage Networking (CRC – 2011).
“Likewise virtualization can enable more virtual machines to safely share a fewer number of physical machines, enabling simpler DR with more flexibility.”
When doing so, however, the DR plan needs to take into account the added complexity brought about by the virtualization layer and the tools necessary to remove or mask that complexity.
“While implementing virtual servers is a great time to rethink how those virtual machines are going to be protected for BC and DR as well as normal HA,” Schulz says.
Keep Regulations in Mind
It is not enough to have a DR plan in place that meets business needs. For many industries those plans must also comply with national or regulations covering the industry.
“Companies are facing increased regulatory requirements and industry standards for business continuity,” says Allan Graham, director of business continuity products for Equinix, Business Continuity Trading Room (BCTR).
“Natural and man-made disasters serve as a constant reminder that risk planning is a vital part of overall IT operational planning.”
Prioritize Servers and Services
Setting up all systems to be fully and instantly recoverable is costly. But not all services have the same value to the business.
“IT professionals need to understand the rank and order of their systems, in relation to their importance to the business,” says Wynn. “IT professionals should evaluate each part of their IT infrastructures and assign each component a value of importance, allowing their recovery plans to enable the most critical services to be available first.”
Don’t Confuse Data Repositories with Full DR Services
Disaster Recovery is a hot topic, and the term is used in marketing services for everything from backup software to fully fledged, manned, replacement data centers. This applies particularly to vendors serving the SMB market. A vendor may be able to restore one client within the promised window, but can it meet the RTO for all customers simultaneously when a hurricane hits?
“Within the US there are literally tens of vendors currently offering ‘disaster recovery’ services, some of which are really just cloud-based repositories who cannot meet top-tier RTO (Recovery Time Objectives) or RPO (Recovery Point Objectives), and/or offer limited SLAs,” says Scott Reynolds, Director of Marketing for nScaled.
“They are, therefore, selling just offsite, platform-constrained, low-performance backup and recall capabilities to SMBs.”
Simple Solutions
There are numerous solutions that cobble together various vendors available for backing up and restoring data, or for running services in the cloud. That may be fine during normal operations, but when a disaster hits it is better to have one vendor to deal with, rather than trying to coordinate the activities of an entire business roundtable.
“With each additional vendor added to the IT portfolio, there is an increased risk of not being able to recover in the event of a disaster,” says Eric Weiss, Vice President of Marketing, Axcient. “This defeats the purpose of using those solutions in the first place.”
Keep in Communication
All of these tips are vital. But in the absence of effective communication during a disaster, they come to naught. If the network is down, how do you talk to your employees and how do you keep your customers informed?
“Most organizations have some sort of central corporate communications system, such as corporate email. But how do you contact your employees when the network goes down?” says Sandra Figueroa, Director of Marketing at Ivycorp, a provider of group messaging solutions.
“And how do you keep from being bombarded by phone calls wanting to understand what is happening. The best solution is to have another means of communicating to this group of people, an alternate network.”
Ivycorp’s Ivytalk can be used as a backup communication network for employees and customers. Using a multi-channel messaging platform that can support two-way communication in both dispatch and chat modes, Ivytalk can be set up in a group/sub-group structure to send and receive messages and maintain the group communication necessary to execute plans and keep everyone informed.