Load Balancer for Production Pega Based applications
Currently customer have pega based applications in production on 7.1.7 on hosted on two nodes running on Windows Platform with Apache Tomcat server. Now they are planning to imlement Pega HA availability and Load Balancer. In this context customer need expertise and advice to proceed with HA features available and LB.
Need help on the below queries.
1. Applications were developed on Pega 6.2 initially and later platform is upgraded to Pega 7.1.7. However still the applications are working fine as per functionality. Is there any limitation if HA features to be implemented as per Pega 7.1.7 and there any challenges involved. Please specify if this is a blocker to implement High Availability Features in Production. 2. Customers have two nodes and if the agents are enabled on both the nodes will there any contention issues. Also specify if Load Balancer is implemented. 3. If user is working on one of the node and other node is down , if the agent enabled on working node will the user request be catered via UI. Please specify if there is any impact if Load Balancer is implemented from this context.
**Updated by Moderator: Vidyaranjan. Removed user added Ask the Expert tag. Apologies for confusion, shouldn't have been an end-user option***
The behaviour of agents have not changed in Pega 7 compared to Pega 6. Agents are background tasks. You can execute the same agent on multiple nodes. If it is a standard agent, then it will pick up one item from the queue and process it. If it is an advanced agent, then the agent activity must handle the sharing of work itself.
1) After impelementing Load balancer settings and If one of the node goes down, will the agents configured on second node run with out any issue or will there be any dependency on other node.
Agents are background tasks and have no dependency on the load balancer. If the agent has been configured to run on all nodes, then it will run on the second node even if the first node goes down.
2) If we configure strandard/ advanced/ legacy agents including AES on both the nodes, will there any contention issues? If contention comes up, is there any best practices to resolve it.
I am not sure I understand here. Are you talking about agents related to AES (in AESRemote ruleset)? I don't believe there will be any contention issues unless your "Advanced" agent is using custom locking of some sort.