From time to time I am getting 499 status codes in the nginx logs of my web/app server (AWS linux AMI + nginx + php).
After reading this other post, I increased the ALB idle timeout from 60 secs to 300 secs.
The problem is still happening and I still can see the 499 errors.
The average latency in the AWS console (ALBs -> Monitoring -> Avg. Latency) is increasing at the time of the 499´s, which is causing the requests to take longer than expecting, and therefore those 499´s happen.
The other problem is that the ALB Healtch Check is also getting a 499 when the problem happens, and that is causing that the ASG to terminate and create new EC2s.
The strange thing is that I am getting the 499´s at the same time in three different environments, that are in 3 different custom VPCs and that are not sharing anything besides the AWS AMI (different RDS, ALB, VPC, EFS...)
I checked several issues from different people in the AWS forums related to this problem:
- https://forums.aws.amazon.com/thread.jspa?messageID=576653򌲍 https://forums.aws.amazon.com/thread.jspa?messageID=287149񆆭 https://forums.aws.amazon.com/thread.jspa?messageID=297948񈯜 https://forums.aws.amazon.com/thread.jspa?messageID=791292󁋼 https://forums.aws.amazon.com/thread.jspa?messageID=291981񇒍 https://forums.aws.amazon.com/thread.jspa?messageID=651803򟈛
How to solve this issue?
