So, you have just lost your most business critical application or worse yet the entire data center. What do you do? Unfortunately I have experienced these exact situations more than I would like but here are ten tips to get you through and more importantly what not to do.

10 Tips for Data Recovery

1. Don’t panic and remain calm! If there is a disaster event the priority number one of any business continuity plan is to ensure the safety of the employees.
2. Evaluate the disaster! The best business continuity plan is the one that doesn’t have to be implemented and being able to quickly evaluate the level of disaster will be critical in making the decision to implement the procedures or not.
3. Communicate! Make regular communication to your team, managers as well as contact your vendors to help determine your options and recruit assistance if needed. Even if there is no status to report, don’t leave anyone guessing or worse yet drawing their own conclusions. This should be done after every other step. Let everyone know the procedures that are being implemented, a conservative estimate of time to complete the recovery and then any update, issue or notice that the business critical systems are available.
4. Know The Plan! Exercising the Business Continuity Plan regularly will help everyone’s familiarity of procedures so there are no questions regarding what to do.
5. Make The Decision! Once you have determined the level of disaster and everyone is safe to operate, it is time to make the decision if you need to implement the business continuity procedures or if the downtime for recovery acceptable.
6. Start the Recovery! Whether one server or two hundred servers, start with recovering the most business critical systems first to restore business operations to a functional level. This should be outlined in step 4 “know the plan” so there shouldn’t be any question which order which servers need to be restored.
7. Where’s the Data? The first step to the recovery is having a duplicate set of data to recover from. This could be anything from archived tape, local disk copy, and a co-location or disaster recovery data center. If it takes you more than 24 hours to receive a duplicate tape from your archive vendor, you might want to start looking for a job now.
8. Which is quicker? The fastest way to recover from any failure is a high availability solution where a standby server (local or remote) and or cluster node sharing a duplicate set of data is ready to go, so starting the services on that server and redirect users if necessary is all that is required. This process can be as little as ten to fifteen minutes to resume critical workload operations.
9. Tier 2 and 3 Servers! These servers are less critical, where archived tape and a greater than 24 hour recovery time objective may be acceptable.
10. Failback Strategy! Once the systems are operational, the disaster is averted and systems are repaired it is time to plan the failback strategy to move the workloads back to where they were originally.

At a high level these are essentially the steps that are taken to recover data in the event of a disaster but there are several technologies utilized in this recovery process. There are several options to consider for backup and recovery and many will depend on the recovery point (RPO) and or time objective (RTO) requirements. Here are the top five most utilized industry standards for backup and recovery.

Top 5 Backup and Recovery Technologies

1. High Availability – whether host based (asynchronous) or disk based (synchronous) replication a real-time copy of the data will be available for a near zero RPO and RTO as little as a few minutes.
2. Continuous Data Protection (CDP) – provides any point in time recovery for restoring previous revisions of specific files. CDP is sometimes required for industry compliance regulations such as, Sarbanes Oxley, HIPAA or SEC, where document version control and retention is required. CDP is a very granular type of recovery process where entire servers can be recovered or just a single e-mail.
3. System State or Server Image Recovery – These systems provide images of a particular server which includes the system state, applications as well as the data and can be kept up to date with real-time replication. The RPO is near zero but because there is a recovery process involved depending on the amount of data that needs to be restored the RTO could be anywhere from 2-4 hours. This is usually a good solution for tier 2 and 3 servers where immediate availability isn’t necessary but the RTO of a tape solution isn’t fast enough.
4. Snapshots – Snapshots are usually configured for every 6 hours and can have upwards of 512 snapshots of a particular volume but that can depend on the hardware vendor as well as what disk space is available. RPO is the interval in which the snapshots are configured and RTO is less than 1 hour.
5. Tape Backup – Typically performed every 24 Hours on a per server basis with a 24hour RPO and potentially upwards of a 24 RTO.

No matter what technology is selected for recovering data the most important items to remember are to stay calm, communicate as often as possible and know the procedures and process in place in order to react as efficiently and effectively as possible to quickly recovery your server workloads and resume business operations to a functional level.