[TriLUG] need url monitoring + pretty graphs

[Aaron S. Joyner]

Ryan Leathers wrote:

>I'm running several instances or Apache in various locations.  I have tools
>to manage these, but frankly, these are not boss friendly.  I need a
>consolidated view of all my sites in one place, with the simple indicators
>of up/down and response time when I knock on the front door of the site.
>Boss types want this in a browsable or emailable graph with pretty colors.
>
>Can anyone suggest an existing tool that monitors several URLs, reports on
>response time, and draws a pretty graph of results for management types to
>gaze at?
>
>I can imagine scripting this up and using mrtg to display the results, but I
>know this has to have been done many many times before.  Any examples to
>share?
>  
>
The best solution I know off-hand for this would be a combination of 
Nagios and RRDtool via something like APAN or one of the many other 
Nagios -> RRDtool interfaces available here: 
http://www.nagios.org/download/extras.php  If you're already running 
Nagios, you're in good shape.  If you're not, you might want to look 
into integrating something similar for performance monitoring into your 
regular uptime monitoring package (OpenNMS, BB, etc).  If your backend 
doesn't easily support this, this might be a great time to change it.  :)

Another option separate from your uptime monitoring solution would be 
something like vanilla MRTG using a few scripts to pull the data you 
need.  These types of scripts are easily googled for - some of the ones 
I've used in the past are the MRTG-PME (Performance Monitoring 
Extensions) which allow you to easily can results from load, disk usage, 
number of running processes, etc into MRTG graphs.  They are 
conveniently available for Solaris, Linux, and BSD.  You might also 
investigate something like smokeping 
(http://people.ee.ethz.ch/~oetiker/webtools/smokeping/ - from the author 
of RRDtool) if you're more concerned about network latency than service 
latency.

Aaron S. Joyner