Location of dependent containers during run

[pagrubel]

Should the images be in the bee_workdir or /tmp or some other temporary directory. Now we have several locations for dependent containers all in <bee_workdir>. We should evaluate if this is what we should be doing. In the past, for Charliecloud it was suggested that they should in a temporary directory, most likely due to the delay in using NFS.

[pagrubel]

We are using redis and neo4j dependencies in containers. The neo4j container is untarred into <bee_workdir>/deps/dependent_container and redis is untarrred to <bee_workdir>/redis_container. I propose that we untar them both into: <bee_workdir>/dependant_containers and name the neo4j container "neo4j_container". This will be better organization and if we need to make it easier to configurize this location.

[pagrubel]

Closing this because it is covered in issue #734