In the original Men in Black comics, an agent named Ecks went rogue and wanted to eliminate the MiB. This little Python library has a similar aim...
Ecks provides a simple way to get data out of a remote machine using SNMP without having to deal with a single MIB or OID.
Ecks is made up of a core class that will collect data via SNMP, and a set of plugins that contain the OID and the code needed to transform the results from nested OID's to usable data.
For far too long people have been abusing their servers by connecting to them using SSH
and then scraping the output of processes like
netstat to find out what their
machines are up to. They do this because it's easier than using SNMP. Ecks now makes the right way the easy way...
The easiest way to install Ecks is:
pip install ecks
You can also clone the repository and run
python setup.py install
The python requirements to use the library are:
- Python 2.6 or higher
Some of the OID's used by the standard plugins shipped with Ecks are not enabled by default on some Linux distributions.
snmpd.conf file should get you enough access to test with:
com2sec notConfigUser default my-secret-community group notConfigGroup v2c notConfigUser view systemview included .1 access notConfigGroup "" any noauth exact systemview none none
If you are worried about other people being able to just read your data then I suggest you restrict the IP address(es) that can access the host as well.
>>> import ecks >>> e = ecks.Ecks() >>> e.get_data('127.0.0.1', 'public', 'disk') [(2, 'Physical memory', 8589934592, 5169360896), (3, 'Swap space', 134213632, 45056), (4, '/', 290984034304, 243201781760)] >>> e.get_data('127.0.0.1', 'public', 'cpu') (21, 9, 68) >>> e.get_data('127.0.0.1', 'public', 'uptime') 18879153 >>> e.get_uptime('127.0.0.1', 'public') 18879153
Note that the last two method calls are interchangable.
This shows how easy it is to get the data out of the system. What you do with it then is up to you...
I currently use this library to feed our internal time series monitoring system and send email alerts when certain thresholds are crossed.
All the hard work is done by the plugins. Currently I have working plugins for:
- CPU Utilization
- Disk Utilization (think
- Load average (think
- Memory Utilization (think
- Network Interface Counters (think
- Process Tree (think
- TCP Connections (think
netstat -an -p tcp)
- Uptime (you guessed it - think
- TCP Stack Statistics
- OS Identification
More plugins coming soon (and more info on how to create your own)...
All base plugins have been tested to work with the default SNMP agent for:
- RedHat Enterprise Linux
- Windows Server
Note: For the windows systems, the load average has no meaning and so does not work.
The way CPU and Memory are reported are different, so use
winmemory instead of
Copyright (c) 2011-2012 Chris Read. See LICENSE for details.