skip to navigation
skip to content

numenta-rogue 1.0.1

Metric collection agent for Numenta Rogue showcase application

Numenta Rogue

Metrics collection agent for the Numenta Rogue showcase application and
consists of two primary components: A long-running metric collection agent,
which periodically polls various metrics and records the results to a local
database, and a separate process for forwarding metrics to a Grok server for

Each metric collection agent is executed in a continuously-running gevent
Greenlet, with a pre-determined wait interval between samples. Each sample
is cached locally in a dedicated RRDTool database. Individual metrics are
implemented as subclasses of `avogadro.agent.AvogadroAgent()`, and must
implement a `collect()` method, which returns the metric value.

The `rogue-forward` process fetches metric data from the individual RRDTool
databases using the `rrdtool fetch` command, and sends the metric data using
the Grok Custom Metric API. Each time the `rogue-forward` process is
executed, the most recent timestamp is cached locally, and is used as a
starting point for subsequent requests.


Install RRDTool::

brew install rrdtool

Install in development mode::

python develop --install-dir=... --script-dir=...

Start key-counter with nohup::

nohup rogue-keycounter > rogue-keys.stdout 2> rogue-keys.stderr < /dev/null &

Make sure that iTerm/Terminal is allowed in System Prefrences > Security & Privacy > Privacy > Accessibility!

Start metric collection agent with nohup::

nohup rogue-agent --prefix=var/db/ --interval=300 --hearbeat=600 > rogue-agent.stdout 2> rogue-agent.stderr < /dev/null &

Forward pending metric data to Grok once::

rogue-forward --server=https://localhost --prefix=var/db

Sample crontab entry::

* * * * * PATH=$PATH:/usr/local/bin PYTHONPATH=... .../rogue-forward --server=... --prefix=... > .../rogue-forward.stdout 2> .../rogue-forward.stderr < /dev/null

Export to local CSV

All metric data is written to a local round-robin database, which only retains
the two most recent weeks of data at any given moment in time. Should you want
the data exported to CSV, use the `rogue-export` utility::

rogue-export --prefix=var/db

The exporter keeps track of the position of a given metric, so you can run
`rogue-export` as frequently as you like, and the `.csv` files in `var/db/`,
will be updated accordingly. i.e. you can periodically sync the round-robin
database to a csv file in `var/db/`



Total CPU utilization as a percentage, as reported by ``psutil.cpu_percent()``.


The percentage memory usage calculated as `(total - available) / total * 100`,
as reported by ``psutil.virtual_memory().percent``


Number of bytes read (total), as reported by


Number of bytes written (total), as reported by


Time spent reading from disk (in milliseconds), as reported by


Time spent writing to disk (in milliseconds), as reported by


Number of bytes sent, as reported by ``psutil.net_io_counters.bytes_sent()``


Number of bytes received, as reported by
File Type Py Version Uploaded on Size
numenta-rogue-1.0.1.tar.gz (md5) Source 2015-08-05 24KB
numenta_rogue-1.0.1-py2-none-any.whl (md5) Python Wheel 2.7 2015-08-05 25KB
numenta_rogue-1.0.1-py2.7.egg (md5) Python Egg 2.7 2015-08-05 24KB
  • Downloads (All Versions):
  • 18 downloads in the last day
  • 65 downloads in the last week
  • 434 downloads in the last month