Justler,
The way that I am currently doing this is to collect information from
the ALERTS. If you dig down into the Smokeping.pm script, you'll see
that at the point where it e-mails an alert, it can be configured to
call a perl script enstead (well, really, anything executable).

So I have a script which writes a line to a file for each offline/online
alert and then I have another script which pulls that information
together into an Excel spreadsheet. It still needs a little work.

The nifty thing is that if you take the UNIX time it went down and
subtract that from the UNIX time it came back up, you have the down
time. Cool.

You could devise the alerts you need for the various international
regions and set your script to execute based on which alert was
triggered.

My next effort is to use the perl bindings to grab the downtime data
from the RRD's themselves using RRD::fetch. But that's going to take a
little more headscratching.

Dan

-----Original Message-----
From: Justler [mailto:[EMAIL PROTECTED] 
Sent: Tuesday, July 22, 2003 2:52 PM
To: [EMAIL PROTECTED]

I need to have the ability to provide a % of uptime.  When the fping
response violates a certain threshold (say 600ms), the line needs to be
counted as totally down.  The response thresholds are different per
region
(international).

What I need to do is calculate a percentage of uptime based off the
actual
no response time, and the violation of my threshold.

How do you guys recommend I do this in smokeping?  Maybe I can't do it
in
smokeping?

--
Unsubscribe
mailto:[EMAIL PROTECTED]
Help        mailto:[EMAIL PROTECTED]
Archive     http://www.ee.ethz.ch/~slist/smokeping-users
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi

--
Unsubscribe mailto:[EMAIL PROTECTED]
Help        mailto:[EMAIL PROTECTED]
Archive     http://www.ee.ethz.ch/~slist/smokeping-users
WebAdmin    http://www.ee.ethz.ch/~slist/lsg2.cgi

Reply via email to