hash out production conda environments

In order to run informatics modules in LIMS, we install the allensdk and other repositories into conda environments on our SAN. The intended workflow is roughly:

1. a change is made to allensdk/internal (or some other repo)
2. a bamboo plan runs tests
3. when those tests pass, a repository maintainer triggers a bamboo deployment which updates the conda environment at a specified path (in practice the update removes and recreates the environment).

The executables table in lims records the path to the appropriate conda environment for each distinct executable.

This system has been put in place multiple times by different people for different projects and repositories. In order to make this situation more straightforward and easy to maintain, we should:

• determine which conda environments are currently in use
• determine which build and deployment plans manage them
• determine who is in charge of each deployment plan
• remove redundant environments
• remove redundant build and deployment plans
• ensure that we have enough build resources for these plans

Start by inventorying the conda environments currently referenced by the LIMS executables table, then map each environment to its Bamboo build and deployment plan and responsible maintainer. Review changes in allensdk/internal and the related repositories only as needed to identify managed environments. Done means the current environments, plans, owners, redundancies, and build-resource needs are documented and the redundant items are removed.