Even all our colelctors were up running, but still some agents failed to be instrumented, and got
Could not register agent with collector as the collector temporarily declined service
Why? Any solution?
Do you have a load balancer between the agents and the collector(s)? If you do, could you bypass it and point the agent directly to the collector? Using collector groups there should not be a big need for a load balancer.
Another idea, can you go to the collector list on one of the agents that do not work:
And make sure that there are no collectors in there that aren't supposed to be there? If there are, stop the JVM, clear them out from the list and then start the JVM again.
Since the other groups installed and configured the agents, I have no access to their conf file. But I checked the agent log and made sure the list of collectors the agent can see is correct.
Might be on that time the collector(s) were running out of resources or in the buffer saturation condition.
Can you check the health of your collectors from the Start Center > Monitor?
Thank you both for answering my qesution! we have 4 collectors grouped together for load-balance. The CPU and memory usages showed on server health looked ok. BTW, what is the buffer saturation condition?
Buffer Saturation: The buffer utilization as a percentage per each connected Collector instance.
Open the Start Center > Monitor > Dynatrace Collector Sizing to also check the CPU and RAM utilization.