[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: free/non-free ladp servers comparizon



David Obadia wrote:
> The performance rates for OpenLDAP compared to Active Directory Server are not
> even close: How objective are this tests?

Read the article. It's fairly clearly stated that the numbers are skewed
based on benchmarking on completely different OS's, which means completely
different setups, installs, etc.

> Search Rate test with one client (Operation/sec):
> Open LDAP on Linux           => 4.5
> Active Directory on Win 2000 => 999
> I'd like to see the tools they've used to make this tests!

Ask Joel. His email's posted right there. (er... see his email, see my
email. I know the guy, he's on the up and up... he's not a MS patsy, if
anything, he's a VMS kind of guy.)

> There is probably caching or something, that completely fake the result...
> Does anyone have a better explanation?

The machine may have had full firewalling on, inspecting each packet...
this would mean that each network transaction would be slower, but internal
operations would be as fast as normal. The RedHat box may have had a generic
kernel, weighed down with all sorts of unneeded overhead. The results may
have been the product of openldap's UI, being command line, and thus not
easy to properly configure. The machine may have been running other services
that the NT boxes weren't running.

There's lots of possibilities, which is why it's not part of the main test.

-Ron

--
Brought to you from boop!, the dual boot Linux/Win95 Compaq Presario 1625
laptop, currently running RedHat 6.1. Your bopping may vary.