Date: Wed, 27 Oct 1999 00:25:01 +0100 From: Aernoudt Bottemanne <bottemanne@capitolonline.nl> To: Stephen McKay <syssgm@detir.qld.gov.au> Cc: freebsd-advocacy@freebsd.org Subject: Re: Need anti-exchange ammunition Message-ID: <3816384C.52ED34B0@capitolonline.nl> References: <199910261348.XAA16538@nymph.detir.qld.gov.au>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi Stephen, Well about costs I can give you some figures. These figures are all from IDG and Gartner Group, so were talking about objective and quality information. This information, amongst others, I use to convine people to go for Linux/FreeBSD/any other Unix: TCO (Total Cost of Ownership): Unix versus NT, in general > 20 % cheaper to have Unix: * users Unix=14,57% cheaper * staff Unix=20,34% cheaper *admin Unix=8.76% cheaper * network Unix=49,77% cheaper These figures are based upon 1000 users in anetwork. When the figures grow, the differences become bigger (in Unix advantage !) Up-time: based on a thorough investigation ( > 100.000 companies...) Unix=23,6 hours per year of the air NT=224,5 hours per year of the air These are the official figures. Several large Dutch organizations, some big banks, calculate 300 hours downtime per year per NT server..... These figures are for Unix on Intel, the downtime of Unix on proprietary hardware is even lower. AIX servers on RS/6000 or SP are leading the compitition here. Good luck ! FreeBSD: with patented Gates-Be-Gone(TM) gets rid of blue screens in a flash ! It forks ! It blits ! Look at those fantastic pixels ! It surfs the web ! You could even host an ISP with it ! Stephen McKay wrote: > First, thanks to those who have responded so far. What I've seen is > enough to show me that there are troubled times ahead. But none of > it will convince the "opposition", for want of a better word. I'm > still hoping for more extreme evidence. > > Forgive me if I don't attribute everyone. You'll be able to spot your > own bits. > > Cost. Normally a strong point, you'd think. But I believe we are in > the position of already having paid. We'd have to un-pay. Very rare > thing for the Government. Leads to finger pointing, and nasty letters > in the paper. So, I think exchange will replace CCMail here, but I'm > hoping to stop it spreading too far. > > The high cost of Exchange and the big iron needed to run it might just > stop the onrushing tide before it engulfs the mail hub, but it is unlikely > to stop it anywhere else. > > In fact, the whole thing has been sold as a money saving operation. But > not from equipment or software costs, but administration costs. For this > sort of thing, management are utterly convinced that point and click is > better than a command line interface. After all, they have no idea what > to do with a command line interface, and their Unix jockeys are way > expensive. Ergo, in with the GUI, out with the CLI! > > To summarise that point (because I'm sure we could have a big advocacy > storm on this one if we really wanted), the reasoning goes like this: > expensive people use CLI, cheap people use GUI, so let's buy a lot of > GUI software. It's not really about sendmail vs exchange. > > Someone asked about sizing. We have approximately the same number of Unix > server admins as Novell server admins, and about twice as many Unix boxes > as Novell boxes. This is partially because the existing mail system is > CCMail, and that keeps them busy repairing it all the time. A couple NT > boxes have appeared under the guise of pilot studies. There are a whole > heap of PC support people who manage desktops and do the user hand holding > stuff. > > Our current mail system supports about 1500 people. We know we could do > the lot with imap and a couple FreeBSD boxes (distributed over the state), > but I've discovered that the killer argument is that the mail program must > come with a scheduler/organiser, just like CCMail. Again, management are > utterly convinced that scheduling and mail are the same stuff. They can > point to CCMail and say "We want one just like this, except that works". > We have nothing to offer, calendar wise. Suggestions solicited! > > The argument about broken Exchange systems being hard to repair because of > embedded SIDs is enough to keep me away, but not compelling at a higher > level. A manager says, "Mail administrator! Fix that box!" and their work > is done. They do that already with CCMail. Nice try though. I'll use it > against some irritating opposition techies. > > Someone mentioned NDS. This is the new wonder product, I'm told. We are > getting it here in a big way, and it will link in with Exchange too. > Apparently it will replace all our account login details on every system > in our entire organisation, and will replace DNS and DHCP. Phew! That's > another story though, and we are well advanced in arguing against it. > Still, if anyone has any reason to believe that NDS distributed replication > doesn't work, I'm all ears. :-) > > In fact, the "Save Our DNS/DHCP" campaign is the only thing to bear fruit > so far. Partially it's because of my catchy slogan: > > "It's a text file. And we *love* it!" > > referring to the plain text config files which we use for everything, and > which we revision control and grep etc, vs the difficulty of doing anything > sensible with a GUI. The slogan is so catchy that I've convinced a layer > or two of management to fight our cause. In the "you win some, you lose > some" situation, I'll keep the DNS, DHCP and password file, and lose the > mail system, if I have to. > > Stephen. > > PS We're dropping in on an Exchange "success story" tomorrow. If you have > any really pointed questions I can ask, feel free to send them in. Ta! > I've got a few myself, and I'll let you know if anything fun happens. > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-advocacy" in the body of the message To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3816384C.52ED34B0>