bac / gypsy-danger

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Pull various Juju KPI data in and report on it

The goal of this project is to provide tools and scripts to help measure the trend of Juju usage, interaction, and growth.

Long running instance data

This script is used to check our API logs for api.jujucharms.com and to judge how many anonymous models are running, how long they've been up for, what versions of juju they're on, and what clouds/regions are they located in.

To get started you need to get the log files onto the system. For this you need credentials to the charmstore logs via swift. These can be obtained by getting them created in enigma via an RT ticket.

Getting setup


  make sysdeps
  make
  source ~/path/to/novarc/for/swift
  source .venv/bin/activate
  make get-logs
  make _initdb
  make longrunning


Loading new logs

Once running you can download the latest log files and update your database with the following:

source ~/path/to/novarc/for/swift source .venv/bin/activate make get-logs make updatedb make longrunning

make longrunning

This triggers the longrunning.py script which will dump all known data out to
the terminal. This is made up of many different queries that can be made
against the log data that's been pulled down and put parsed with `_initdb`.

    long-running/longrunning.py run summary
    long-running/longrunning.py run latest-summary
    long-running/longrunning.py run model-ages
    long-running/longrunning.py run models-per-day
    long-running/longrunning.py run latest-versions
    long-running/longrunning.py run latest-clouds
    long-running/longrunning.py run latest-clouds-regions


- summary: how many models have we seen in the db
- latest-summary: for the latest day, how many models were seen
- model-ages: for all model uuids found how many days old are they
- models-per-day: for each day how many models were seen on each day
- latest-versions: for the latest day, how many models are on which versions of Juju
- latest-clouds: for the latest day, what clouds were those models on
- latest-clouds-regions: latest-clouds with a breakdown per region of the cloud

About


Languages

Language:Python 86.8%Language:Makefile 9.1%Language:Shell 4.1%