I think C. External monitoring tools or log aggregators like Splunk must be configured to recognize new node. Two ways to configure that in mulesoft. Either through log appender setting for log4j, or configure Splunk agent to pull logs from server. If Splunk agents are configured to pull logs and other data from nodes, agent will need to be configured to pull logs from new node as well. Cannot be D because firewall rules configuration are between API clients (I am assuming end user HTTP request generator) and load balancer. Nodes are not visible to API Clients. Cannot be B because existing load balancer can be reconfigured to add new node.

in question it is mentioned all rquest are throuh LB [Mule applications deployed to this D, is correct, cluster are invoked by API clients through a load balancer.] , but not mentioned anywhere that external monitering tool is configured, for montering Anypoint control plane is there, so Anser is D

C. External monitoring tools or log aggregators must be configured to recognize the new nodes

D can not be right. Client ---> Node is not correct. Client -->LB is correct. I go with C

On a separate note, is this also valid? "A new load balancer entry must be configured to allow traffic to the new nodes"

Option A is in no way the right answer. For API client requests to reach a new node, a firewall opening is needed from LB to new node. Read option D carefully, it says "accommodate communication between client and node". That doesn't mean that API clients are sending the request directly to nodes.

D is not correct as there is no need to enable the firewall settings between api clients and nodes. C is correct

C. External monitoring tools or log aggregators must be configured to recognize the new nodes

Im sure is C…. clients hit the LB so they don’t care … you should not have to update any implementation everytime new nodes are added.. that is on the LB layer.. and the firewall rules are between the LB and the workers… the only option that make sense is C that is external to mule.

D Load balancer is already in place, but it needs to be configured to also consider new nodes.

I think it should be C

Answer should be B. Persistent object store in on-premise is created in shared memory of cluster. I believe it wont have any impact if another node is added. Hence not option A.