If you are subscribed to this blog via RSS, you may have noticed that May and June have been two empty months in terms of writing, and tobe honest the entire 2019 has not been so prolific as usual. This is because I worked, and I’m still working, on some large projects that took a big chunk of my time. I’m still writing these days, but the outcome is coming out in big pieces instead of weekly posts. The first one is this, about Veeam Availability Console.
The new VAC book!
I liked Veeam Availability Console (VAC) since the first day I saw it (and it was still called Managed Backup Portal). It has an insane potential to build and sell managed services, and I’m sure we are just scratching the surface of what’s possible. It all started as a manager for remote locations full of Windows Agents, but obviously when you give a tool in the hands of end users, they will find out new use cases by themselves that we didn’t even thought about. That’s what happened for example with remote Veeam Backup & Replication servers: we started to see providers using VAC only to manage multiple VBR servers, without any agent involved.
What we realized over time, with all these deployments and the critical services that providers were offering on top if it, was that some better architecture was needed. At the beginning, VAC was deployed as a full installation inside a single machine. That was good enough for some fairly large deployments, even in the thousands of managed objects, but sometimes it’s not just about size: one single machine may be able to control large estates, but it has little to no redundancy when it comes to be up and running as much as possible.
So, we started to think about different deployment scenarios that are not about how to use VAC, but about how to build VAC infrastructures!
Scale-out
I’ve always been a fan of scale-out deployments, and I explained why it makes so much sense in a dedicated post I wrote a long ago: The future of storage is Scale Out. It was about storage, but the concepts there can be applied to any software system. And that’s what I did also for VAC infrastructure: I splitted the different components into dedicated machines, and studied how each of them could have been deployed into multiple and redundant copies. The final design is like this:
But since the world is never black or white, or say binary, I also took into account the idea that a provider may not want to go all-in at Day1, and want to start small and then grow as his business grows. So, the book really starts from the simple deployment, and then explains how to seamlessly migrate to more complex deployments, by splitting and replicating the different components.
In addition, you will also find chapters to learn how to manage the backend SQL database, the single place where all VAC data is stored, and the SSL certificates, all automated via Let’s Encrypt technology, even when multiple DNS names are used by the VAC resellers.
There are still chapters that need to be developed and written, since a reference architecture is a ongoing and never ending work. There’s always something to improve, new information to add, new feedback from the field that can change what we said before.
For this reason, exactly like my previous book about Veeam Cloud Connect, I dediced to make it an online and live book, by usign markdown as the writing language, and Github Pages to host the book itself. So, to read it, just go here: https://dellock6.github.io/vac-book/ and let me know in the comments here on in the Veeam forums what do you think about it.