Skip to content

bashwork/speleo

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

71 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Speleo

What is it?

The aim of speleo is to be as good or better than splunk at a much cheaper prices (read free).

Why?

Because splunk is expensive and every attempt to clone it or to make a logging platform has not been successful in my view because of:

  • it just isn't as pretty or engaging
  • it just isn't as easy to use
  • it just won't scale to the needed size
  • because I wanted to write this for kicks and grins

What's inside?

So I am putting all the hard work on the shoulders of giants:

  • log agent

    I using flume for now, but whatever can send messages to the collection service can be used.

  • log collector

    I using flume for now, but whatever can accept messages from the log agent can be used.

  • log propagation

    I am using a plugin for flume to send the messages to everywhere they need to go: hadoop, elastic search, etc.

  • log parsing

    The idea here is to train a bayesian classifier to decide what kind of message we have just received (json, syslog, xml, etc). Based on that, we parse it before sending it to the index.

  • log indexer

    I am using elastic search for now, but maybe we can abstract so any of the indexers can be used (solr cloud for example).

  • service api

    This will be a django (or jetty) api used to manage everything else: statistics, users, permissions, saved searches, etc.

    This will also need to move cold log data out of the cluster based on defined policy (for now we are leaving it as a manual step).

  • service storage

    This will be abstracted using SqlAlchemy (or the java equivalent) and will be used to persist data used by the api.

  • long term log storage

    This will be hadoop and the interface to this will simply be a linked in version of hue (or any other tool the user feels like using).

    Storing off log information will be handled manually by users for now, however, we can add administrative tools to help in the future.

  • speleo frontend

    That is this! Ideally a rich web application that communicates with all parts of the service via rest api's providing a seamless experience regardless what device they are on (computer, tablet, mobile).

What else?

I dunno, I'll post this when I figure it out.

License?

How about the BSD License, I find that fair.

About

splunk, just not as expensive

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published