Closed pbelmann closed 2 years ago
Hi, The image is indeed large due to use of taxonomy database inside. Other option is to have databases folder mounted in the docker image to avoid long initialization inside the container each time you boot the image. We will create a lite image. What is the reason for a smaller image size use case?
Well in my case the system I'm using for running my pipeline has just 20GB of disk space on the root partition. If run a pipeline with multiple docker images, then there is not enough space left. I agree that mounting your databases (which are in my case on a different partition) to a container is the way to go. Thanks! I'm looking forward to your lightweight image.
Hello,
Lite image of MOB-Suite v3.0.3 with tag kbessonov/mob_suite:3.0.3_lite
without initialized databases is available on the DockerHub resource. You can pull it (docker pull kbessonov/mob_suite:3.0.3_lite
) and run it docker run --rm -v $(path2databases_dir):/mnt/databases mob_recon -i assembly.fasta -o outdir_name --database_directory /mnt/databases
. Note: make sure you mount the directory with databases so that mob_init
routine would not be triggered causing database install inside the container.
Awesome! Thanks!
Hi,
thank you for this tool! Your docker image with 4.71GBs of size is quite large. I guess the reason for this is that your databases are part of the image.
Is there any way that you provide a second docker image that does not download all databases?