graphite-spark
A sparkline for graphite data: â–„ â–„ â–† â–‡ â–„ â–„ â–ƒ â–ƒ
When to use
Its 3am you have just ssh'd to a server and due to performance problems. The Load seems high, but you look after 2000 servers... Is this a normal load? You need to know what is ''high'' for this system?
A quick graphitespark -l and you can see a 24 hour history of the loadavg for this system.
You know now that is a normal loadavg and you can investigate something else.
Usage
usage: graphitespark [-h] [-f filesystem] [-l] [-c custom graphite path]
Display a sparkline of this or other servers
optional arguments:
-h, --help show this help message and exit
-f filesystem, --filesystem filesystem
-l, --loadavg
-c custom graphite path, --custom custom graphite path
Installation
The argparse module is a required dependency. If not already installed:
% pip install argparse
To install graphite-spark system wide:
% git clone https://github.com/daniellawrence/graphite-spark.git
% cd graphite-spark
% python setup.py install
Then, just execute:
% /usr/bin/graphitespark
To run standalone without installation:
% git clone https://github.com/daniellawrence/graphite-spark.git
% cd graphite-spark
% PYTHONPATH=src bin/graphitespark
Example - Check Load average
Checking the load averages of the current system that you are logged into straight from the command line!
server 1 % graphitespark -l
-24hours@1h blocks
1min load avg.
â–ˆ â–… â–† â–„ â–„ â–† â–† â–„ â–ƒ â–ƒ â–ƒ â–‚ â–ƒ â–ƒ â–‚ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–„ â–… â–ƒ â–ƒ
Max: 2.95 Min: 0.6 First: 2.95 Last: 0.82
-24hours@1h blocks
5min load avg.
â–ˆ â–„ â–„ â–„ â–„ â–† â–‡ â–„ â–„ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ
Max: 2.26 Min: 0.54 First: 2.26 Last: 0.76
-24hours@1h blocks
15min load avg.
â–ˆ â–† â–„ â–„ â–… â–‡ â–‡ â–† â–„ â–„ â–„ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ â–ƒ
Max: 1.82 Min: 0.53 First: 1.82 Last: 0.59
Example - filesystem capacity
Checking the the filesystem capacity of a local filesystem on the current host, from the data in a remote graphite server.
server 1 % graphitespark -f /var/tmp
Filesystem capacity: /var
-24hours@1h blocks
â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ â–ˆ
Max: 11.0 Min: 11.0 First: 11.0 Last: 11.0
As you can see the script was able to work out the mount point of the filesystem so that the data was still found in the graphite system.
Example - custom graphs
Checking anything that you want from any system that you are logged into. As the data is in the remote graphite system.
server 1 % graphitespark -c 'systems.s.server2.loadavg.15min'
-24hours@1h blocks
â–… â–… â–… â–„ â–ˆ â–ˆ â–… â–„ â–‚ â–„ â–ƒ â–‚ â–‚ â–‚ â–‚ â–‚ â–‚ â–‚ â–ƒ â–ƒ â–ƒ â–‚ â–ƒ â–‚ â–‚
Max: 15.03 Min: 1.73 First: 9.07 Last: 1.73
