Weird Stuff I Saw While… Working with Multi-Subnet Availability Groups
Details
Our presenter this month is Rick Lowe, Founder of Data FLowe Solutions and MCM!
Session Details: Despite the fact that SQL Server Availability Groups are built on top of a stack of rather complex technology, over the years things have gotten to the point where getting an AG up and running for a simple two-node HA setup is not as intimidating for most DBAs as it may have been in the early days. When multiple subnets get involved, however, things can get weird quickly. Some stories that will be told include:
· Inability to bring AG online at DR site after shutting down main site
· Seemingly random and unpredictable connection timeouts suddenly occurring a month after AG is extended to a second subnet
· Failure of the primary replica after rebooting the secondary during patching windows
We're review this, along with discussion of configuration adjustments that can help keep some of these issues at bay!