SYNBIOCHEM Galaxy server that supports the design pipeline infrastructure. Pablo Carbonell, SYNBIOCHEM, 2020.
The system is built from two docker images:
sbcglxbased on the Galaxy Docker Image.sbc1based on docker-annaconda.
The reason for creating two separate images is to have an independent sbc1 image with all required libraries for running the SBC tools. In order to be able to run sbc1 from sbcglx, the latter has to share the docker socket from the parent and have the permissions to read and write:
-v /var/run/docker.sock:/var/run/docker.sock:rw
Create account galaxy with following id details:
uid=1450(galaxy) gid=1450(galaxy) groups=1450(galaxy),27(sudo),999(docker)
Main scripts:
build.shbuild the images and install the SBC repositories.start.shwill start the galaxy server.stop.shwill stop and remove the container of the server.
Other scripts:
tweak.shcontains some commands that might be required to run at least once in order to install the Galaxy server in a new computer.
A container running an nginx server is used in order to employ virtual domains.
The Galaxy tools are in https://github.com/synbiochem/galaxytools. When possible, the tools are run within the sbc1 container by Galaxy. For some apps, there were some issues and I have written directly the docker invocation in the tool. In the future, this should be unified and made more automated.
Another issue is that some jobs that connect to ICE like plasmidGenie or plasmidWriter, etc use multithreading and the docker container might continue running if the job is killed from Galaxy. A better way of managing these jobs is necessary.