jmeter-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Philippe Mouawad <>
Subject Re: Performance difference between released 2.6 and nighly r1095243
Date Wed, 29 Feb 2012 15:11:02 GMT
This setting is true by default so it works the same as old versions.
And if working fine it's job is to reuse the SSL context providing better
performances than set to false.

So I don't think it's the explanation unless it was tweaked to false in
your current configuration.

What makes you think it's the cause of issues ?


On Wed, Feb 29, 2012 at 3:59 PM, ivan.rancati <>wrote:

> hi Philippe,
> I think I have found the difference between the nightly build and the
> released 2.6
> While running with the gc switches that you suggested, I found this line in
> the jmeter.log of the released 2.6
> jmeter.protocol.http.sampler.HTTPSamplerBase: Reuse SSL session context on
> subsequent iterations: true
> Nightly: if there is an authorization manager, it will attempt basic
> authentication only on requests where the server sends a challenge
> Released 2.6: if there is an authorization manager, it will always attempt
> basic authentication. The api that I am testing is not sending any
> username/password.
> Now, I think that entering a Bugzilla issue is probably not required.
> However, I did not see a mention of this change on
>, so it's possible that other users
> will find the same problem.
> I think the setting in that changed is
> #https.use.cached.ssl.context=true
> If I get it right, the proper solution, if one initial ssl http call is
> required, is to put the authorization manager only in the relevant Thread
> Group, otherwise it will have a global effect (i.e. also on http calls
> where
> the server does not challenge).
> Or am I missing something obvious?
> thanks
> Ivan
> --
> View this message in context:
> Sent from the JMeter - User mailing list archive at
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

Philippe Mouawad.

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message