10-26-2016 09:09 AM
Our Brocade vTM (virtual Traffic Manager) appears to have a memory leak to the tune of ~500MB a day which we are monitoring closely.
My question is, in order to avoid an IP failover during the day ie when free memory runs out, I would like to manually failover the elastic IP used by the Traffic Group (containing 2 AWS Ec2 instances) during a maintenance window.
Can anyone confirm the best way of doing this please? I have tried "interrupting network traffic" ie stopping the active instance in AWS as well as switching the elastic IP association manually in the AWS console but neither of these lead to a smooth failover.
Other suggestions are to run stop-zeus on the active node within $ZEUS-HOME which I am yet to try.
Is there a recommended or tried and tested way that anyone knows of - I was looking at the REST API and thinking that possibly the TrafficIPGroups.setPassiveMachine or TrafficIPGroups.setTrafficManager commands might be of use...?