Skip to content

This script uses the sampling facilities built into Qumulo Core to provide an overview of capacity consumption first by user, and then by area of the tree.

License

Notifications You must be signed in to change notification settings

Qumulo/capacity-by-user

Repository files navigation

Deprecation Notice

This software will shortly be deprecated and archived. If you have any issues, please reach out to Michael Kade directly.

Capacity By User

What's the point?

This script uses the sampling facilities built into Qumulo Core to provide an overview of capacity consumption first by user, and then by area/path of the filesystem tree. It reports the file owner whether it's a local Qumulo admin, NFS user, or AD/LDAP user managed externally. It has the added feature of being able to assign a dollar amount to capacity to report the capacity utilization in money terms in addition to bytes consumed.

Scenario

Imagine you have a hundred or a thousand artists or researchers all busily collaborating on a project. They create data, and sometimes they forget to go clean it up. You want to get your army of talented professionals to reduce their capacity consumption. What do you tell them? This script breaks down a tree by user and then shows, for each user, where that user consumes space in the tree.

Example commands

python3 capacity_by_user.py -C qumulo.local -U admin -P $PASS /home -s 10000 -x 3 -D 50 --fancy-output

This command will look at a sampled set of items from the "/home" directory and subdirectories (which in this case is mostly users' scratch space). It assigns a value of $50 to the cost per terabyte (maybe that's the amount you bill your users per month). It will limit the output to a maximum of 3 leaves per user to keep things succinct.

Dollar capacity by user output on the command line with top 5 users shown

python3 capacity_by_user.py -C qumulo.local -U admin -P $PASS /home -s 10000 -x 3 --fancy-output

This command is the same as the first example. It reports by capacity rather than dollar amounts as the -D 50 argument was removed.

Byte capacity by user output on the command line with top 5 users shown

Usage

Requirements

  • python 3.4.11 or newer
  • python qumulo_api package 4.0.0 or newer
usage: capacity_by_user.py [-h] -U USER [-P PASSWORD] [-C CLUSTER] [-p PORT] [-s SAMPLES] [-c CONCURRENCY] [-m MIN_SAMPLES] [-x MAX_LEAVES] [-D DOLLARS_PER_TERABYTE] [-i] [-A] [--fancy-output] path

positional arguments:
  path                  Filesystem path to sample

optional arguments:
  -h, --help            show this help message and exit
  -U USER, --user USER  The user to connect as (default: None)
  -P PASSWORD, --password PASSWORD
                        The password for the specified user. If not specified, you will be prompted later.
  -C CLUSTER, --cluster CLUSTER
                        The hostname of the cluster to connect to (default: qumulo)
  -p PORT, --port PORT  The port to connect to (default: 8000)
  -s SAMPLES, --samples SAMPLES
                        The number of samples to take (default: 2000)
  -c CONCURRENCY, --concurrency CONCURRENCY
                        The number of threads to query with (default: 10)
  -m MIN_SAMPLES, --min-samples MIN_SAMPLES
                        The minimum number of samples to show at a leaf in output (default: 5)
  -x MAX_LEAVES, --max-leaves MAX_LEAVES
                        The maximum number of leaves to show per user (default: 30)
  -D DOLLARS_PER_TERABYTE, --dollars-per-terabyte DOLLARS_PER_TERABYTE
                        Show capacity in dollars. Set conversion factor in $/TB/month
  -i, --confidence-interval
                        Show 95% confidence intervals
  -A, --allow-self-signed-server
                        Silently connect to self-signed servers
  --fancy-output        Organize each output item into a 'pruned' path tree

About

This script uses the sampling facilities built into Qumulo Core to provide an overview of capacity consumption first by user, and then by area of the tree.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published