Can your backup survive a hurricane?
It’s that time of year again, hurricanes, tornadoes and severe thunderstorms are back again. Which means you’re at risk, especially if you haven’t done a recovery test of your backups recently. What are you waiting for, a natural disaster?
To ensure that your recovery is smooth there are a few things you should keep in mind. Including making sure you have a backup, that you know how to recover and that you have a system to recover to if yours is no longer usable. If you don’t have these things then you’re going to be in trouble when disaster strikes.
Store Backups offsite
While it may be common sense that you need to keep a copy of your backups offsite for recoverability, we encounter many companies who are not religious about removing tapes from the drive. It’s imperative that you store your backups offsite in case of a disaster. It doesn’t matter if you use physical tapes or virtual tapes, you need to make sure there is a copy in a second location.
One issue with offsite backup is how accessible it is following the disaster. For example, if you keep your backup in a vault at a bank and the bank was also impacted, you might have an issue. Alternatively, if you store it with Iron Mountain, it may take some time to get to you following bad weather. Make sure your backups are offsite and accessible. Test the process of getting the backups returned to you so you know how long this will take.
Recovery reports accessible
Doing a restore is not hard; it’s making sure you have the right tapes to restore and that you restore them in the correct order that is more difficult. You do your backups every day, but do you know what is on the tapes? Do you know which tapes you need to rebuild your environment back to a close to the point of failure as possible? If you don’t do a full system save every night then you will have to piece your backup together from the tapes you need. Without a recovery report, how will you do this?
If you use BRMS, you have a recovery report. This report is the key to piece your system back together. We recommend emailing a copy to yourself, so you can easily access it in an outage. Alternatively, send a copy to your business partner so that they can help you in a disaster.
If your site is destroyed and you can’t reload your environment, what is your plan B? Do you have a DR system in a second location that you can restore your backups on? Do you have a contract with a cloud provider and need to tell them you’re sending your tapes, or do you need to find someone who can help you? It’s important to know what you will do if you need access to another system.
Planning and Testing
The middle of a natural disaster is not the time to prove your disaster recovery plan works. You need to have a plan in the event of a disaster, but just as important is testing that plan before disaster strikes to be sure that it works. You don’t want to find out that it takes you three days to get your media and then find out that you’re missing pieces you need, while the business is breathing down your neck to get things back online. Then only to find that you missed a critical tape and now you are back to square one. It’s better to find out what can get in the way when there is no pressure. So, test your DR plan and test it often. Practice makes perfect.
A Real Customer Example
We have a customer in South Carolina, which is prime real estate for hurricanes. They asked iTech about a DR solution to be able to keep the business running when a hurricane was hitting South Carolina.
iTech recommended our Disaster Recovery on Demand service which provides a “skinny LPAR” for 11 months of the year, and a full LPAR for one month of the year for testing purposes. The service includes having iTech perform the restore of the customers backup and then provides the customer with access to test for 30 days. DR on Demand seemed like the right solution, until the first hurricane warning.
The issue that the customer faced was that they had to ship the most recent backup to the data center, they wanted to have the most recent data, so they want to do a backup before they shut down the system. Getting the backup tape to the iInTheCloud Data Center became the next challenge. If you don’t get the tape out before the airlines stop flying (because they don’t like to fly into hurricanes), then you can’t recover. Then there was the issue of the delivery time once the tape arrived in Michigan. Not to mention that restoring from tape takes time as well. The recovery time became an obvious issue.
To solve the issue of getting the backup to the data center promptly, we recommended a VTL solution. Using a VTL reduces tape handling, and it also offers replication. VTL’s make it easy to get a copy of your backups offsite, but it requires a second VTL in a second location.
iTech offers a VTL replication service which allows customers to replicate their backups from their local VTL to iTech’s VTL in the cloud, which means you only have to invest in one VTL device. The VTL’s use deduplication to reduce the amount of data that is replicated every day, so the replication is fast once the first backup is seeded.
The customer invested in a VTL and replicates their backups to iTech’s VTL daily. The next time that the customer wanted to test the recovery process, it was a breeze. iTech expanded the LPAR to have the resouces needed for testing and restored the backups. Everything was available in less than 4 hours, and the customer is very pleased with their recovery time.
This can be you
You can be like the customer above and have a tried and tested disaster recovery plan. Contact email@example.com if you want to discuss how we can help you with your DR strategy and testing.