Scan all files under $DIR into the database (or update existing records under $DIR):

$ dusqlite scan $DIR

Print a summary of disk usage scanned into the database:

$ dusqlite report
Tags:
    umdata                    11.7 gb    65266
    conda                      8.8 gb   263338
Scanned Paths:
    /home/562/saw562
        saw562   w35           1.4 gb    18179
        saw562   w48           0.0 gb        4
        saw562   S.U           0.0 gb       29
    /short/w35/saw562
        saw562   w35        1278.4 gb   659444
        saw562   w48         158.2 gb     2737
        hxw599   w48           6.4 gb       16
    mdss://w35/saw562
        saw562   w35         305.3 gb        6

Find files under $DIR:

$ dusqlite find --older_than 1y --group w35 $DIR

Run a check on a path:

$ dusqlite check directory-group-readable $DIR

dusqlite reads a yaml config file from $HOME/.config/dusqlite.yaml to set the database path etc.

To see the current configuration run:

dusqlite print-config

You can tag directories to add summaries to dusqlite report. This is done through the Configuration, a config file to create the report above is:

tags:
    umdata:
        paths:
          - /short/w35/saw562/UM_ROUTDIR
          - /short/w35/saw562/cylc-run
    conda:
        paths:
          - /short/w35/saw562/conda

The meaning of tags is up to the user - they could help monitor the size of temporary directories, or allow you to find files related to a specific project more easily.

dusqlite can be extended to support other types of storage, such as tape stores, provided there is a way to get directory information (e.g. by running ls -lir)

To scan a tape store pass a URL as the path to scan, which specifies the storage system and path. Supported URL types are:

• NCI MDSS: mdss://$PROJECT/$PATH

• Add check reports
• Handle multiple paths in cli arguments
• Checks idea: temporary files of failed NCO/CDO commands
• Add more find arguments, e.g. size, mode