Starting today, new Vapor deployments will ensure a separate AWS Lambda function exists for running your queued jobs.
Before this change, queued jobs executed within the CLI Lambda. This Lambda is also used for running deployment commands, CLI commands, and scheduled jobs. However, there were potential problems with this approach. For example, if queued jobs were consuming all of the Lambda's capacity, no other CLI invocations would be able to process, causing deployment commands and scheduled jobs to fail.
Beginning today, we recommend that you set the
queue-timeout configurations for every environment. Being able to control the
queue-concurrency separately from
cli-concurrency ensures your queues won't consume all of the available capacity and cause other invocations to fail.
Here’s what we've been up to recently.