I’LL NEVER FORGET THE DAY I told a client we weren’t going to be able to boot their backup. I was relieved that at least we had their data, but incredibly frustrated that the image could not be virtualized on demand as our sales pitch had promised. I believe it weighed into the client’s decision not to renew our contract the following year. Shortly after that we vowed to never sell another vendor’s BDR in a box.
Backup is all about trust. Your clients are trusting you to recover their data and systems when disaster strikes. They are counting on you to bring your “A” game. They trust that you have done your homework, know how the BDR solution works, and that you have an intimate knowledge of their software and hardware. They trust that you are testing the backup, and that it’s recoverable in appropriate recovery point and recovery time objectives (RPO/RTO).
We learned two important lessons from that failure. First, we realized that we needed to have more regular, in-depth conversations with our SMB clients to define their RPO/RTO expectations. The root cause of the BDR failure was that the product we were using wasn’t designed to support files as large as our client had grown into producing, preventing timely virtual boot. Had the vendor alerted us to this issue, we both would have had an up-sell opportunity.
Second, we learned that it is a strategic disadvantage to outsource your core competency. My clients trusted me to keep their data safe and recoverable.
We quickly made the decision to pull the plug on the BDR vendor and build and manage our own backup solution. Turned out it was good timing too, as that vendor became unresponsive anyway.
There are lots of reasons why building your own BDR solution might make sense for your business as well. Perhaps you’ve been let down by a backup provider as we were. Perhaps you want to increase your margins or differentiate your offering from the commoditized BDR space. Or maybe you are not satisfied with the available data center locations.
Before you go down this path, though, let me spend a moment to talk you out of it. This is a mature market, with lots of solid companies to choose from. Building your own solution is going to require a great deal of time and a substantial up-front financial commitment. However, the ROI comes in higher profits and more control.
Our philosophy is to treat all business data as important. Therefore, we believe that all servers and workstations on the network should be part of the backup strategy. We surveyed our client base and found that the majority were not interested in spending significantly more for a local boot-ready server. What they wanted was dependable backup to the cloud, with a fast way to recover to the original device.
As an underlying technology, we use StorageCraft ShadowProtect for physical and virtual servers and Veeam for workstations. We also use a Qnap 653A NAS device to store copies of the local backup images, and we replicate those images out to our data center where a monster version of the same NAS lives. Each client’s data is encrypted during backup; it remains encrypted both during transport and at rest, on both ends.
We rent all the hardware equipment and software, eliminating customers’ up-front costs. Initially, we worked with leasing companies to finance this, but eventually we built up the cash flow to self-fund the equipment purchases. While that may sound frightening, a solid contract and good client vetting and business practices will keep you out of a jam.
If clients want to step up to a bootable BDRU (as we call it) for a faster RTO, then we add in a server sized to their data set today plus expected growth over the next 36 months. We also rent the server, and because we are only storing the number of boot images specific to the client, we can size the device more appropriately.