We are doing our application development in cloud environment version 7.2, we have 4 nodes for our environment.
I know that when one of the node is over burdened in cluster with heavy DB pool size or if there are any memory issues pega will automatically swap out that node and replace new node in its place and when this new node is added the agent schedule rule is formed based on agent rule.
my question here is when a environment faces this issue multiple times, will pega swap one of the running node with a old node that was disabled earlier if it does is the agent schedule rule formed again based on agent rule or it will pick the old rule.
for example if initially environment was running on nodes 1,2,3,4 and due to some issues 2 and 4 were disabled in that place 5 and 6 were added now again due to some issue if 1 or 3 fail will 2 or 4 nodes brought back into the running status or not and how their agent schedule rule is formed
and what kind of algorithm is used in backend for switching or swapping between nodes like: roundrobbin or something
The Cloud instances run on VM (virtual machine). The replacement of the node in Pega Cloud is purely managed by ASG (Auto scaling group) and triggered by ELB (Elastic Load Balancer). Everytime ASG will bring up a new instance in response to an event, such as OOM or hardware failure etc.
For your question on whether Pega Cloud ASG will bring a old node to swap, the answer is NO - it will always be a new node from the VM pool.
Pega Cloud load balancing uses round robin routing algorithm.
Please let me know if this answers your question ?