I remember that report. At first they wouldn't tell me when I called who paid for the tests. It was only after a week of bitching how they screwed up their tests that I finally received the answer.
There were things they did in their detailed analysis that proves they boosted NT's default settings and limited Samba. They did the same things a couple years back with NetWare 5.0. For example in NetWare they turned off read ahead cache but boosted the tcp windows to around 17k (default is around 8k I believe). In short I don't put much stock in mindcraft when they intentionally do things like this. Hell... I even went as far as to speak with a Novell Engineer about the settings. Mindcraft was "claiming" they had received the settings from Novell itself. I've been a Novell CNE much MUCH longer than I've been a UNIX Admin and can tell that they are soo full of crap it's unbelievable. Look at KeyLabs Inc for truely independant testing. Sure they take money from companies like Novell and Microsoft however I used to work there years ago and they didn't try crap like this. Frank >Lewi wrote: > >>> >>> I have a job to replace Netware to linux as linux server on my >> >division, > >>> >>> well, when I'm trying to search information in internet, I found that >> >linux with samba is slower than NT > >>> try a look at: >>> http://www.mindcraft.com/whitepapers/openbench1.html >>> I know that the result maybe can't be compared today >>> does anyone know anything about this, and what do u think? >>> >>> it is right to choose combination between linux and samba? any >> >suggestion? > >>> and maybe u can give me an url about information that related to my >> >problem? > >>> >>> thank you >>> >>> -- >>> ichtus >>> ------ >>> Lewi Supranata .K >>> ICQ: 50643061 >> _______________________________________________ Redhat-list mailing list [EMAIL PROTECTED] https://listman.redhat.com/mailman/listinfo/redhat-list
