I would have thought that buildpack compilation failure was an uncommon scenario. However just today there were 185 stopped containers sitting around that seem to be from buildpack compilation failures.
These are never cleaned up and require manual intervention if we don't want them to build up.
The app:docker:delete_created_status_containers rake task clears up stopped containers from quite an early stage - these don't take up much space, they're just annoying to have around.
However there's a whole other group of containers from the next compilation phase and these can end up taking 10s of GBs. You can clean these with the more robust app:docker:delete_stopped_containers rake task.
The question is, where do these containers come from? Why are there so many of them? And how can we minimise them so manual intervention isn't required?
I would have thought that buildpack compilation failure was an uncommon scenario. However just today there were 185 stopped containers sitting around that seem to be from buildpack compilation failures.
These are never cleaned up and require manual intervention if we don't want them to build up.
The
app:docker:delete_created_status_containersrake task clears up stopped containers from quite an early stage - these don't take up much space, they're just annoying to have around.However there's a whole other group of containers from the next compilation phase and these can end up taking 10s of GBs. You can clean these with the more robust
app:docker:delete_stopped_containersrake task.The question is, where do these containers come from? Why are there so many of them? And how can we minimise them so manual intervention isn't required?