World Backup Day is a great time to remember all of those times #whenbackupsgowrong. That’s what we thought when we recorded the World Backup Day edition of the Druva Podcast. Between myself, Stephen Manley, and Dan Frith, we have 75 years of experience with backups and restores, so we’ve got quite the collection of interesting stories.
Our collection of backup stories
We talked about the event that launched my career, which was when I was unable to restore an Oracle database, which happened to be the purchasing database of a $35B credit card company. That was the day I learned that Oracle needed to be put into a special mode during backups. That’s one lesson I’ll never forget.
Dan Frith regaled us with a story about an email archive system that stored its data on a single EMC Centerra (Remember those?). The data on the Centerra was deleted in a very efficient way, and the backup tapes came to the rescue. Like a lot of tape restore, it took a long time, but they did eventually get the data back. They also learned the lesson of how long a restore like that would take and decided to buy a second Centerra and sync them together.
Stephen’s story was also of a very different time, one where Yahoo! ran the IT world. For those of you that grew up in the world of Google, there was a time that Yahoo! was at the center of the universe. They were a huge client for a lot of companies, and Stephen worked for NetApp at the time, who provided the filers that stored Yahoo! Mail. Stephen explained how they were using Network Data Management Protocol (NDMP) to do their backups and it was taking forever, mainly because the filers were very busy during the backups and NDMP was designed to deprioritize itself (i.e., nice in the Unix world) under other processes. This did remind me of one of the many reasons why Druva chose not to use NDMP to back up filers, despite the fact it is the “official” way to do so.
Towards the end of the podcast we talked about how we could help a potential backup administrator be more interested in the job, and there were a few ideas. Use a service so the job isn’t so hard, use a backup product that supports analytics and let it manage that, and finally use the service as a central source for business knowledge, because doing their job properly requires interfacing with all of IT.
Your backups shouldn’t be an afterthought
In this unsure time of the COVID-19 Coronavirus, it’s good to see some things continue to happen normally. Druva’s customers continue to have their backups work flawlessly because they don’t need to manage them. Once our customers set the policies and schedules, the infrastructure part (the hard part) is all up to Druva. We then make it all happen via Amazon Web Services’ (AWS) lights-out datacenters and their various cloud services.
Don’t be like me, Stephen, and Dan. Train your backup people well, have replicated copies (not tape) for critical data, and make sure your backup system isn’t designed as an afterthought. It should be part of the initial design so that you’re not making any design decisions that will make it harder to back up.
Happy World Backup Day!
Druva is here to help your business stay resilient, protected, and compliant while your employees work remotely. Learn more about our free offering.