As part of my CICD deployment, there is a volume my_volume
that gets created on docker-compose build/up
, that needs deleting every deployment.
Therefore the CICID script calls docker volume rm my_volume
before docker-compose build/up
.
But if a build fails, subsequent builds will error out on docker volume rm my_volume
, because the volume doesn't exist.
How can I remove this volume only if it exists?
CodePudding user response:
In order to ignore failure while calling "docker volume rm my_volume" use the below order-
set e docker volume rm my_volume docker-compose build/up true
CodePudding user response:
You can ignore the errors:
docker volume rm ${some_volume} || true
Or you can start by making sure the project is completely down:
docker-compose down -v
docker-compose up -d
Or you can start labeling your volumes and containers when testing, and prune those labels:
docker volume prune -f --filter 'label=ci-test'