A new Harvest EC2 instance has been created y'day using the AWS cloud formation template. All the docker containers that this cluster manages are in restarting mode with "unable to load template" and "template not found on confpath" errors. I will send an excerpt from the logs on email.
#All docker containers on a EC2 Harvest host are in "restarting" state
glass brook
hallow cargo
received
glass brook
Hi @hallow cargo , so carrying on our conversation here - rather than copying the /opt/harvest/conf from a container which was needless because there isn't a running container yet. I just removed the 2 lines referring to the volume mounts from the manage_harvest. Will that not solve the problem ? I ask because the docker-inspect still shows references to the volume mount and the containers are still failing
hallow cargo
that's the correct change to make but the container needs to be stopped like so ansible-playbook manage_harvest.yml --tags stop then restart with ansible-playbook manage_harvest.yml --tags upgrade
glass brook
yes of course I have rerun the manage_harvest to re-create the containers
glass brook
yes
hallow cargo
if you do that, and a stop and upgrade, you should be all set
glass brook
Thnaks @hallow cargo that worked. We usually prefer to run the "manage_harvest" as an ansible pipeline from an external source, but I had to break my rules to run it locally with the tags as you said to get this sorted. All ok now 🙂
Another thing that I wanted to ask, Chris, was if there is any way for us to retain old data aftre renaming the cluster in the Harvest.yml file. I have just tested that and found that a new docker instance gets created and any old data on the systems gets lost