Matt That IT Guy

Racking And Stacking In A Cloud-Based World


A look at Veeam Availability Orchestrator

One of the greatest benefits that I have received with the Veeam Vanguard program is the people. Being able to chat with some of the brightest folks in the industry is something that is hard to put a price on. That’s why I have come to really look forward to Vanguard Day before VeeamON. This is the time where we get some great presentations and great conversations going.

Kicking the day off was a great presentation from Michael White, who is heading up the Veeam Availability Orchestrator project. To get us all in the right mindset, he did a great presentation on the nuances of Disaster Recovery vs. Business Continuity. What was great about this presentation is that Michael, who has years of relevant experience, was able to give us real-world examples of issues that he has run across in the field.


VAO FeaturesWith the foundation set, we jumped right into VAO. So, what exactly does VAO do? To start, it is VMware-only at the moment and its goal is to handle all of the ‘stuff’ that goes with performing a DR / BC plan and automate as much as you can. For example, it can build out your documentation for you and hand-off a Word file that you can do with what you please. This is also useful for finding changes over time, whether or not they were planned or not. VAO will also handle automated tests – think about building out plans based on SureBackup where you can be sure that your DR will work as you planned.

That’s all cool, but how do we control how all of this stuff works? As Michael pointed out, triggering a DR failover can actually cause a denial of service during the failover process. What if you have a malicious employee who wanted to take down production, how could you ensure that they won’t trigger an event. Conversely what if you need the “OK” from multiple folks to trigger the event? VAO has you covered with baked-in security. You’ll be able to grant permissions based on roles, but you’ll also be able to do something like “at least two members of this group must agree to initiate a failover”. Another example given was that you might be able to spin up a Virtual Lab and only allow access to specific users. Although not the intended use case, this might be a great way to delegate permissions to developers to create test environments.

We had a glimpse at the install process, which is fairly straightforward (next, next, next, finish), however, it might take a while to install due to the complexity of the software. Minimum requirements for a PoC is 2vCPU and 8 GB of RAM for 10 replicas. Configuration can include tasks such as defining a SQL Server, setting up SMTP, adding Veeam Backup & Replication servers, etc. Also installed is a special version of Veeam ONE, that offers more than the free version, but is different from the “standalone version”. This version will be included with the licensing for VAO.


Once you get your setup configured, it’s time to start making plans. These will be the actions that are performed when a plan is kicked off, so tasks such as building out your VM groups (e.g. bring up DCs and then Exchange). Knowing your workflows will be critical to building out plans. You won’t want to just lump all of your VMs into one job, but rather you’ll want to break them into smaller chunks. The best part about building these jobs? You can use vSphere tags. This will allow you to tag your VMs based on whatever SLA you want (or dependencies, etc) and have them automatically added to your VAO job.

Logging is actually noteworthy as it will even log things like PowerShell error messages (from customer provided scripts), or invalid login credentials. I’ve been driven nuts in the past by trying to troubleshoot software failures with little to no clues as to what is failing, or why. Of note is that PowerShell is the only scripting language supported at the moment.

VAO is still very much a work in progress. The Veeam Vanguards were first told about the product in June of 2016, and since then it has gone from a concept to a running demo. Based on what I saw, the product has come a long way. I foresee installation being a challenge for a lot of organizations, not due to technical issues, but due to business processes. Getting IT folks talking to business folks isn’t always a quick conversation. All too often I have seen businesses try to automate a process, only to realize that they don’t actually have a formal process in place. A project like this, however, can help remedy that.

VAO is slated for a late 2017 or early 2018 release.

Disclaimer: I was invited to VeeamON by Veeam. All of my expenses, including food, transportation, and lodging are being covered by Veeam. I did not receive any compensation to write this post, nor was I requested to write this post. Anything written above was on my own accord.

One thought on “A look at Veeam Availability Orchestrator

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.