All Collections
FAQ
Disaster Recovery
Disaster Recovery

How we keep your data safe across our systems

Mikel Lindsaar avatar
Written by Mikel Lindsaar
Updated over a week ago

One thing that is often overlooked when selecting a software provider like MetaPulse is finding out how the data you are storing on the servers is backed up and protected.

We have already two articles on this:

These go over how we protect your data and keep it backed up so that it is safe in multiple locations.

But another issue that can arise is what would happen if our servers failed completely? For example, what if our servers crash, or the data center they are running on fails? In this case, the backups are basically useless as there is no server to run them on.

Many SaaS companies have setup their servers that run in the cloud by manually connecting into the server and configuring each system bit by bit over a period of time. This is useful when starting out, but is dangerous in an ongoing sense, because if that server is deleted, then the company would have to try and recreate that server, remembering all the 1,001 little steps they did to get the server running, all while their customers are screaming for the service to be restored.

At MetaPulse, we do this another way.

Every server is built using a computer program. That computer program is updated every time we want to make change on the servers and the computer program is then run in order to make that change.

Additionally, every time we deploy a new release of code, we are actually building an entirely new server every time (using the previous one as a template) and then putting multiple, live copies of that server into production.

Every time we apply a security patch to the server itself, this program is updated again.

This program is stored in a completely separate location to the servers that are running the MetaPulse platform. This allows us to recover from a data center failure by just creating a server in a new data center and then running the program to configure the new server and get it up and running ready to restore our data and get you access again.

We also run whole copies of our system in multiple locations around the world. We have copies running on the East and West coast of the USA as well as one in Australia, allowing us to rapidly restore data into a live running system if that was needed, instead of having to wait to rebuild the entire stack of servers required to run MetaPulse.

Recovery Time Objective (RTO)

As all of our code, both for the platform itself and the code needed to build the platform is stored in multiple locations and systems, our RTO is 48 hours.

Recovery Point Objective (RPO)

Due to the way our data is backed up (both via hourly snapshots and daily offsite backups) our RPO is hourly with regards to our databases which are backed up hourly via snapshots and then daily through a backup hosted on a secured S3 bucket.

Summary

So, not only is your data backed up. But our data we need to build the servers to provide you with your data is backed up and available to us to rebuild our entire system at a moments notice.

It's really just another way that we take our role in the custody of your data seriously.

Did this answer your question?