Session reliability and stability is our highest priority.
Our Collaborate team constantly monitors through both automation and human control the good performance and health of our infrastructure.
When a specific server is showing early signs of a potentially degraded experience, we pro-actively take it out of service, so it doesn't get to a point where it would crash and seriously impact the users experience for a longer period.
When a server is taken out of service :
- The disruption to users is short, but it does require them to rejoin the session manually. In most cases they can rejoin within seconds, in some cases it may take several minutes.
- It affects a very small percentage of sessions, hosted in that particular server out of the several hundreds of servers we're running.
- It doesn't affect all of your institution : your institution's session are spread out across several servers in your region
This is not a service outage, as the affected users are able to rejoin shortly, and the very vast majority of users (on other servers) don't experience any interruption.
We've decided to share on status.blackboard.com when a server is taken out of service, so that if a user tells you they've been temporarily disconnected, you can first check if the timing matches this "server taken out of service" and if it does it very likely is the cause, and if it doesn't you should go ahead and create a ticket so we can investigate for a different cause.