Actions
Issue #7485
closedDocumentation: Pulp all-in-one container with CentOS7's docker
Start date:
Due date:
Estimated time:
Severity:
2. Medium
Version:
Platform Release:
OS:
Triaged:
Yes
Groomed:
No
Sprint Candidate:
No
Tags:
Documentation
Sprint:
Quarter:
Q4-2020
Description
This is not really a bug, merely a documentation issue.
The all-in-one-container recipe populates the image's /var/run
(a symlink to /run
) and expects its contents to be available at container run time. The docker implementation that is provided e.g. by CentOS7 however mounts a tmpfs on /run
, effectively hiding the original container's contents. This causes the container to fail early after startup, like this
FATAL: could not create lock file "/var/run/postgresql/.s.PGSQL.5432.lock": No such file or directory
The workaround is to specify an additional /run
volume, suppressing this behaviour of the docker runtime. Docker will copy the image's contents to that volume and the container will start as expected.
Note that e.g. RPMs from the docker community repo are not affected by this.
Actions