We are receiving Http 5XX and 4XX errors from the ELB. I have configured the ELB access logs in athena and analyse the same but doesn’t got any success. Is there any way to get the RCA for these errors.
Any pointers to troubleshoot this issue will be appreciated.
What would be the default idle timeout for the ELB? On further analysis to 5XX errors, we found that we are receiving 502 & 503 from the backend instances.
Also, in every backend instance there is a CPU spike after every 20 mins. And we have any other task/ job scheduled in the back end instance. Initially I thought this was because of the process that push logs to S3 which takes the CPU, but this is not the case as the process runs on an hourly interval while the CPU spike is on 20 min interval.
Any thoughts ??
Anyone can provide the headers for the tomcat application raw logs??