Many times we'd like rever jobs to run after something happens where that something could take a lot of time (longer than a laptop power cycle). For instance, after a conda-forge version bump PR has been merged I'd like to make a new version of my docker container and upload it somewhere. A naive approach would just to have the terminal wait for the PR to be merged, but that terminal could be shutdown sometime and then the job would be lost.
A potentially better approach would be to register amazon lambdas for these jobs so that they run on always on infrastructure.
Many times we'd like rever jobs to run after something happens where that something could take a lot of time (longer than a laptop power cycle). For instance, after a conda-forge version bump PR has been merged I'd like to make a new version of my docker container and upload it somewhere. A naive approach would just to have the terminal wait for the PR to be merged, but that terminal could be shutdown sometime and then the job would be lost.
A potentially better approach would be to register amazon lambdas for these jobs so that they run on always on infrastructure.