I have an AWS ALB that sends requests to AWS Lambda.
Normally the request_processing_time is less than 0.01 second and target_processing_time less than 0.1 second. However, occasionally the request_processing_time can take more than 2 seconds. As this only happens once every few weeks, it is quite difficult to reproduce the long latency case.
What might be the root cause? How can I prevent this from happening again?