Watching the news late last August, we saw that Hurricane Harvey was headed for the Texas coast. Abtech has customers in our StorTrust backup and disaster recovery cloud that would be right in the path of the hurricane, especially one of them, a supply company to the petrochemical industries.
We contacted our customer to talk about preparing for a possible disaster recovery invocation. As a result of the call, we decided to do some advanced preparation and set up their disaster recovery environment. Using Rapid Recovery’s virtual standby server features, we created images of each server required for the disaster recovery. As Rapid Recovery takes each recovery point in their production environment, it is replicated to the StorTrust cloud and the virtual standby server feature applies that data to the images of their servers. This allows a very small recovery point objective and minimizes lost data should the worst happen.
Our StorTrust cloud service includes an annual test. During this test, we configure and spin-up the customer’s disaster recovery environment and set up the networking to connect everything. Once the test is finished, we store the configuration for use during a disaster. In preparation for the possible invocation, we applied the configuration to the virtual standby servers we created and we were ready to spin-up if necessary. We supplied a unique IP address to the customer should they have to access their servers.
On the day Hurricane Harvey hit Houston, we sat near our phones waiting for a call. 18 hours into the event, we finally got it. Our customer asked us to start spinning up their environment – they had lost communications with their servers in the colocation facility. Because the StorTrust cloud does not rely on websites or menus to start a disaster invocation, the one call was all our customer had to make to get things started. In his case, that was important because they did not have a network connection to work from and would not have been able to set up their environment and start it.
We started to spin-up the servers and make them active. We were almost ready to put everything on air when received another call saying that network contact with their servers had been restored and asked us to stand down. It turned out that one of their colo’s network feeds had flooded and they were able to fail over to another feed entering the building from a different direction, which is proper for a good colo facility.
Our customer finished riding out the storm without further incident and we put their environment back in normal backup mode. Having everything ready ahead of time in this instance allowed our customer’s personnel to worry about their families instead of how to keep the company running.