Nrthugu

I have had a problem for a few days with EVERY EC2 INSTANCE that I create.

I set up the group security, the jira installation file, the database. Then I open the installation URL and a DB test connection is working but when I submit, it takes a few minutes loading then the whole server is down and the installation is aborted.

BTW: I'm new in AWS so I only changed and added ports in the security group.

Here are some screenshots:

• Connection DB


• Submit


• Instance down


• Inbound traffic


• Outbound traffic

The most likely cause is you're out of RAM. Jira is quite high on requirements, so if you're trying to use an instance with, say, 1 Gb of RAM, it won't work, and the processes will get killed with OOM (out of memory) exception.

Another approach would be to create an instance with ECS-optimized AMI (https://docs.aws.amazon.com/AmazonECS/latest/developerguide/ecs-optimized_AMI.html) and run Jira as a Docker container (https://community.atlassian.com/t5/Jira-articles/How-to-run-Jira-in-a-docker-container/ba-p/752697).

I’ve seen this occur using the Atlassian CloudFormation template. The issue is that the database setup takes so long that the load balancer pings eventually time out, so the load balancer decides that the node is down and removes it from load. You can confirm if this is the case by looking into the load balancer status on the AWS console.

There are probably better ways to handle this, but I got around it by accessing the Jira node directly to perform the setup (ie. going around the node balancer and hitting port 8080 directly). You may need to setup a bastion host and forward connections over SSH to do this.