Hobbit monitoring help needed 🔗 link

Hobbit (4.2.0-with patches) is running http-tests for a particular
system. This system is very unstable and Hobbit proves it, but now and
then the http-check for that host fails in a strange way.

The page output turns to YELLOW for 10 minutes(!) while each of the 5
tested URL's fail (status RED):

[snip]
Wed Jan 9 21:25:04 2008: Server timeout ; Server timeout ; Server
timeout ; Server timeout ; Server timeoutSeconds:

 http://some.url/ - Server timeout

Seconds:    10.06
[snip]

Now 1 minute later, this same test turns into RED instead of yellow.
Why first yellow, then red? And then all alarms are triggered at
once?!

Anyway, I do not understand Hobbit's response?!

Any help much appreciated,  Peter

On Wed, Jan 09, 2008 at 10:35:49PM +0100, Peter Welter wrote:

Do you have a "badhttp" definition for that host (in bb-hosts) ?


Henrik

Yes, I do, but I somehow forgot... But that explains a lot, thanks!

I think I will remove this tag. It's like the example badhttp:1:2:4.,
but it is still so often RED (more than 4 failures, is it not?) and
then Hobbit still alerts me at night :-(

So I use the Hobbit-script-feature to restart the websserver and use
STOP afterwards; no more pager alerts; just restart this "fine" piece
of Tridion-software :-p

Thanks again,
Peter

2008/1/9, Henrik Stoerner <user-ce4a2c883f75@xymon.invalid>: