A Practical guide to StatsD/Graphite monitoring

In addition to this helpful guide, note that statsd &#x2F; graphite both spring some unfortunate surprises on new users, e.g., graphite changing your data across retention rates and time scales [0], graphite changing your data at different plot widths (?!) [1], statsd believing that only count and time data deserve to be aggregated [2], etc.<p>I have no alternative to suggest, however. Perhaps Cube [3], but unclear if it has any user community.<p>[0] <a href="http://stackoverflow.com/questions/10820119/graphite-is-not-graphing-anything-for-ranges-bigger-than-7-hours" rel="nofollow">http:&#x2F;&#x2F;stackoverflow.com&#x2F;questions&#x2F;10820119&#x2F;graphite-is-not-...</a> [1] <a href="http://graphite.readthedocs.org/en/1.0/functions.html#graphite.render.functions.cumulative" rel="nofollow">http:&#x2F;&#x2F;graphite.readthedocs.org&#x2F;en&#x2F;1.0&#x2F;functions.html#graphi...</a> [2] <a href="https://github.com/etsy/statsd/issues/98" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;etsy&#x2F;statsd&#x2F;issues&#x2F;98</a> [3] <a href="https://github.com/square/cube" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;square&#x2F;cube</a>

This is a very nice article.<p>One nitpick- You don&#x27;t need to use statsd as an intermediary in order for your application to send metrics via UDP; just set ENABLE_UDP_LISTENER to True in carbon.conf and graphite will accept metrics on UDP itself. Other options are TCP(obviously) and AMQP.<p>I love how simple Graphite&#x27;s plaintext protocol is; it&#x27;s nothing more than a line of text with &lt;metric path&gt; &lt;metric value&gt; &lt;metric timestamp&gt;. This has lead lots of software to integrate graphite support and makes it easy to do yourself. In a pinch I&#x27;ve even set up a cronjob reading a value from &#x2F;proc and sending it to graphite via netcat.<p>Graphite shines at generating graphs, but it&#x27;s ability to return JSON is also very useful. For example, I&#x27;ve written a script (<a href="https://github.com/sciurus/grallect" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;sciurus&#x2F;grallect</a>) that plugs into Nagios and generates alerts based on system metrics sent by Collectd to Graphite.<p>My two frustrations with graphite-<p>You have to choose a single aggregation method. I&#x27;d like to be able to store the average, minimum, and maximum values.<p>Sometimes I find it hard to query for the data I want. E.G. To check the percentage of space used on each filesystem I have to fetch example.com.df-<i>.df_complex-used and example.com.df-</i>.df_complex-free separately and calculate the percentages myself because asPercent(example.df-*.df_complex-{used,free}) would combine all the filesystems.

I use a statsd compatible alternative called statsite.[1]<p>It&#x27;s written in pure c and behaves like you would expect statsd to, with some additional improvements. I&#x27;m definitely more comfortable deploying it as opposed to installing and managing a node.js application.<p>[1] <a href="https://github.com/armon/statsite" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;armon&#x2F;statsite</a>

Graphite committer &#x2F; co-maintainer here...<p>Feel free to point out useful things graphite could do better (constructively only) and&#x2F;or some of your favorite posts or tools used with graphite. We aren&#x27;t too far off from 2 quite massive releases (0.9.11 &#x2F; 0.10) and are thinking about departing from some of the legacy bits moving forward. I&#x27;m looking at you python 2.4

Everyone considering using Graphite needs to take a look at Librato Metrics (<a href="https://librato.com" rel="nofollow">https:&#x2F;&#x2F;librato.com</a>).<p>It&#x27;s very affordable and dramatically simplifies management of your stats. I am so glad to put Graphite behind me.

Excellent guide. We (at <a href="http://www.hostedgraphite.com" rel="nofollow">http:&#x2F;&#x2F;www.hostedgraphite.com</a>) will be pointing new users towards it and I&#x27;m sure many will benefit. Thanks!

37Signals wrote a piece on how to instrument your Rails app: <a href="http://37signals.com/svn/posts/3091-pssst-your-rails-application-has-a-secret-to-tell-you" rel="nofollow">http:&#x2F;&#x2F;37signals.com&#x2F;svn&#x2F;posts&#x2F;3091-pssst-your-rails-applica...</a>

I created a Vagrant VM for StatsD and Graphite. Maybe it&#x27;ll be of help to someone <a href="https://github.com/jpadilla/statsd-graphite-vm" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;jpadilla&#x2F;statsd-graphite-vm</a>

Thank you. This is awesome.<p>If you have the same metrics coming in from multiple app servers, how can they be viewed in aggregate and separately?

Graphite &amp; Carbon was an absolute nightmare to get installed on CentOS -- we had to settle installing it on Ubuntu instead.