Reaching out to see if anyone has successfully put the puppet openshift bsn components with the ha enhancements into a fault tolerant environment?
I've been using it for some months but recently had some time to run more tests on it. I'm noticing shutting off one of the nodes (mongo and activemq) leads to a 1/3 chance for a user operation to fail. Eg. Refreshing the web console will call every 3 attempts leaving a code 500 error for the user.
Haven't had the time to see if data is being restored after the node is brought back up etc. On my to do list..
Is this something anyone's noticed? Configuration issue?