PHP workers in AWS EBS

PHP

1. messages are sent to SQS.

2. A static amount (configurable) of PHP workers are kept alive by the Supervisor.

   • Given the nature of PHP, we optimize workers to stop and restart every X minutes.

3. While workers are running, they attempt to fetch messages from SQS.

   • If a worker gets a message then it starts working on that job and takes it to completion.

   • If a worker crashes while processing then the Supervisor will bring it back up.

     • What will happen to the message that was pulled off the queue by that "crashed" worker?

     • Will the resurrected worker start work on the same exact message again?