My questions are related to HA blueprints. Suppose I have a 2 master/1 worker setup on AWS/Azure. Now one of my master is down/faulty and I want to add a new master(master3) to existing cluster.
Also, I want to remove the faulty/unhealthy master(master1) from the cluster.
While instantiating the cluster, I had used the HA blueprint with 2 hostgroups (master1 and master2) with their set of components/services.
Now if I say spawn master3 node on cloud platform which has all the required accesses to the existing cluster. Will the following approach work for getting a HA enabled healthy cluster with new master3?
Will Ambari take care of the installing and starting the components/services and also rebalance the cluster with HA settings?
Is there anything other approach using which I can achieve the HA again? OR
Will the re-configuration of all the components in HA be handled outside the ambari's scope?
Please let me know your views. I have limited exposure to ambari at this time so feel free to correct me or atleast give some pointers.
As the blueprints focus heavily on the host_groups concept which usually means some highly specialized master setups and then a more generic worker model, I feel that using blueprints beyond the initial cluster layout really works best for when you are adding more workers. My INITIAL recommendation (I'm arm-chair QBing this w/o all the details) would be to simple go to Ambari's UI and add a new host via the wizard process and then assign whatever master and worker processes you need.
Thanks for the reply @Lester Martin. Will the blueprint take care of setting up/re-configure component level HA with this new node?
In my scenario, I am setting up cluster using the HDP's ansible repo. I plan to add new node using the same mechanism where I can get the VMs in the existing cluster with active master (as user's input) and make sure this new node is added to the cluster via Ansible