Understanding Recovery Time Object (RTO) and Recovery Point Object (RPO) is vital for disaster recovery…
June 18, 2020
What is RTO and RPO ?
Hello. Welcome back to another Portworx Lightboard Session. My name is Ryan Wallner, Technical Advocate here at Portworx. Today we’re gonna be talking about RPO and RTO. So RPO, R-P-O, Recovery Point Object, and RTO, Recovery Time Object. They both relate to a disaster timeline and the disaster recovery plan that covers it. We’re gonna cover Recovery Point Object first. Now, let’s say we have an application here that incurs a disaster somewhere along this timeline and we have a latest backup somewhere around here. Now, this little timeline here, this little gap, represents our RPO and this also represents data loss. And what I mean by that is, if your backup is every two hours and our disaster occurs one hour after that, you have about one hour of data loss. Now, if our recovery happens over here, so our application is back up and running, again, this part of our timeline is our RTO and this translates to downtime for your application and, essentially, your users. So your application might be down for, say, 20 minutes, per se. Now, Recover Point Object is one of those things that, depending on your application, you’ll have to define what’s appropriate for your application.
If you’re in healthcare or financial, you might want a zero RPO plan. And what this means is, your downtime and your data loss has to be zero or near zero. So, best case scenario, your timeline is right here and your RTO also is right here. And so, the way to achieve this is having a replication strategy for your disaster recovery plan to allow data to be always available in both your primary and your DR site. Now, other applications may have a way to kind of restore data from a cold or metadata type of thing, and one hour might be totally reasonable. Again, missed event in the application. And as far as RTO goes, this also depends on your procedures and your automation for doing application backup. In Kubernetes, the scheduler defines when an application gets started and how long it takes to come up. So, the best case scenario is your data’s available in your DR site running Kubernetes and the scheduler turns on an application. So it’s a matter of seconds to minutes before your application comes up in that site. So again, keep in mind both RTO and RPO and have a disaster recovery plan in place with the correct data management tools underneath to support you. Until next time, thanks for watching.