Performance testing Kerberos

Sam Hartman hartmans at MIT.EDU
Thu Jul 8 13:57:11 EDT 2004


>>>>> "dougyoungwith" == dougyoungwith  <dougyoungwith at northwesternmutual.com> writes:

    dougyoungwith> Is anyone aware of Kerberos being tested from a
    dougyoungwith> performance testing standpoint?  My company is
    dougyoungwith> doing a proof of concept on a third party version
    dougyoungwith> of Kerberos and would like to set some baseline
    dougyoungwith> parameters.  I use Mercury's Loadrunner for
    dougyoungwith> testing and am unable to find any information about
    dougyoungwith> Kerberos on the Mercury website.  Any input
    dougyoungwith> would be greatly appreciated.   
    dougyoungwith> Thanks  

I'd recommend setting your performance targets based on your needs and
application rather than on other Kerberos test results.   

It's also not clear which aspects of Kerberos performance your
environment cares about.  Do you care about KDC authentications per
second?  What about application level authentications per second?
What about GSSAPI confidentiality throughput?

--Sam



More information about the Kerberos mailing list