Rocky week in the news cycle – I’m trying not to follow or worry about that too much. It’s always mostly negative news. First day of Spring occurred, we had a couple of really nice early-Spring days, it was good.
Work was more of the same – meetings upon meetings. I’ve been half-joking tracking the total hours of scheduled meetings (including all the double / triple / ++ booked meetings). 63 hours of scheduled meetings for the week, averaging 12.6 per day this week. The rolling average so far for the calendar year is 50.375 hours per week and 11.16 hours per day. It’s….not ideal.
That said, I did manage to carve out time to tinker and learn. There’s an emphasis right now at work around disaster recovery planning, and one of the thing I know I wanted to tinker with was building an active/active service on top of VMs.
When I think active/active, I think of service availability that can withstand regional outages. In Azure, for example, it would mean picking two Azure regions that are geographically separated with enough distance. The definition of “enough distance” will almost certainly differ from org to org.
In my tests, which I hope to have a writeup for in the next few days, I’m leveraging Azure Load Balancers and Traffic Manager Profiles to support the global load balancing. The setup of the environment was actually pretty simple. The service I’m hosting is merely a static web page (I’m not a dev), but that doesn’t really matter when trying to prove out resiliency. Full writeup soon on that….