The /var/lib/docker/volumes/SImA_cluster_fs volume is used to temporarily store the output of batch analysis jobs, e.g. results files, etc. Data is stored in that volume only temporarily, and SImA does make an attempt to clean the volume periodically; if the folder size exceeds a pre-configured maximum size the internal slurm cluster uses a cronjob to clean up the cluster job folders in /cluster/jobs/.
By default the pre-configured maximum size is set to 100 GB, this value is customizable. This is defined by the CLUSTER_RESULT_DIR_MAX_SIZE variable which can be added to the custom.env file which can be found inside the SImA installation directory.
e.g.
# services/custom.env
CLUSTER_RESULT_DIR_MAX_SIZE=10000 # in MB
It is safe to perform a manual clean-up of the SImA_cluster_fs volume as long as you do not target any running batch analysis jobs.
The below example command would delete sub-directories (one sub-directory is created for each plate analysed) which are greater than 30 days old, meaning any currently running batch analysis job would not be impacted. The sub-directories contain results files which have already been returned to SImA and would be available for users to access via the UI, hence no critical data would be touched.
$ find /var/lib/docker/volumes/SImA_cluster_fs/_data/jobs -maxdepth 1 -type d -mtime +30 -exec rm -rf {} \;