[hobbit] bbnet-test https bug

[Brian Majeska]

Hi Henrik,

On Mon, Dec 1, 2008 at 1:18 PM, Henrik Størner <henrik at hswn.dk> wrote:

> In <18b05330811191559g2713c930r11e8334fc395ef35 at mail.gmail.com> "Brian
> Majeska" <brian at majeska.net> writes:
> >environment: Hobbit 4.2 with allinone patch on RHEL5.1
> >It appears that using the https tag and defining a port introduces
> >50-200ms of latency to the results of all bbnet-tests.
> >I had the following tag working fine for months in bb-hosts:
> >cont=kvm;https://server.domain.com/dsview/protected/login.do;Avocent
> >The need came up to move this particular webserver to a non-standard ssl
> port:
> >cont=kvm;https://server.domain.com:9443/dsview/protected/login.do;Avocent
> >As soon as I did this 'all' TCP graphs, http, ldap, smtp, etc... had a
> >new plateau 50-200ms higher.  I ran bbnet-test with the --debug and
> >--timing options but didn't see anything out of the ordinary, then
> >again I don't know what I should be looking for that would help to
> >resolve it.  Let me know what you want and I'll provide it.
> >I was able to duplicate the problem on a separate dev hobbit server as
> well.
>
> I have no idea what could cause this, but I dont see anything in the
> Hobbit code that could account for it. The port number for each test
> is determined in advance, before the tests begin. So the codepath
> used to perform a network connection does not depend on the test
> using a default or non-default portnumber.
>
> Is there any kind of firewall between your Hobbit server and the
> system(s) you're testing ?
>

No, there is no firewall between the boxes.

I was able to duplicate the problem using a basic apache/mod_ssl setup to
using port: 9443 to rule out the possibility of the Avocent/jeti webserver
causing problems.

I baselined the test with the following tag in bb-hosts:
192.168.1.210 server01.domain.com       #  http://server01.domain.com

Then added the https tag:
192.168.1.210 server01.domain.com       #  http://server01.domain.com
https://server01.domain.com:9443/

And observed the same 50+ plateau jump on the graph for the original test.

--
Brian Majeska
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.xymon.com/pipermail/xymon/attachments/20081203/39b8f2f2/attachment.html>