Force RockOn cleanup in postgres

magicalyak · September 29, 2017, 1:51pm

Have we considering using a script to force a cleanup in postgres in storageadmin for rockons, dcontainer, dcontainerenv, dport, dimage, etc with a big warning? This would be very beneficial for anyone with a container stuck removing or other strange scenario.

If someone manually removes the container outside of rockstor, this would be needed, otherwise the UI acts as if it’s still installing or removing the rockon.
For testing custom rockons, this is also beneficial
using a script in /opt/rockstor/bin may be the best path also

Ideally, a flag could toggle an extra option to do this in the UI and perhaps a script, like the debug mode, could trigger it?

Just a thought…

phillxnet · September 29, 2017, 4:18pm

@magicalyak Hello again, good ideas.

Have you had a look at the currently undocumented:

github.com

rockstor/rockstor-core/blob/master/src/rockstor/scripts/rockon_delete.py#L26-L36


DOCKER = '/usr/bin/docker'




# Console script to delete containers, images and rockstor-db metadata of a
# Rock-on.
def delete_rockon():
try:
    name = sys.argv[1]
except IndexError:
    sys.exit('Delete metadata, containers and images of a '
             'Rock-on\n\tUsage: %s <rockon name>' % sys.argv[0])

It was added to ‘force’ remove ‘problematic’ named rock-ons. One at a time, and might be a good place to start with extending such ‘emergency’ maintenance tools re rock-ons.

phillxnet · September 29, 2017, 4:53pm

@magicalyak Just tracked down the issue this script was added against.

and the associated commit:

Hope that helps.

magicalyak · September 29, 2017, 6:24pm

Yes, it’s the same issue, you guys already thought of this