SAS Viya CAS Worker nodes migration on AWS

[nikhilreddy]

We are planning to migrate our CAS Worker Nodes to larger instance, I have couple of questions

1. While spinning up new instance for cas worker nodes we will setup same hostname as before, do we also need to maintain same IP address or hostname should be good?
2. As of now all our servers – microservice, cas controller, worker1 and worker2 are on AWS cloud within same placement group and while migrating the cas worker to another instance family within the same placement group we don’t have the instance family we needed and facing capacity issue. So AWS is recommending to spin up the server outside of placement group.

How would be the impact if we spin up the cas workers away from placement group i.e away from cas controller and microservices ( controller and microservices are staying in same placement group). is that considerable ?

If we Spin up the workers outside of placement there would be some  latency in communication when compared to servers within placement group but still we have the server. Not sure if we can go with this approach. Any suggestions would be helpful.

Thanks in advance,
Nikhil.

[alexal]

@nikhilreddy ,

Each target machine in your deployment must have a static IP address and hostname. There is constant communication among the CAS controller and the worker nodes, so you have to reduce delays in the communication as much as you can.