Date: Thu, 25 Mar 1999 18:33:13 -0600 (CST) From: Christopher Palmer <cpalmer@jig.ordway.org> To: "Scott I. Remick" <scott@computeralt.com> Cc: FreeBSD Questions <freebsd-questions@FreeBSD.ORG> Subject: Re: Out of Swap Space hangs system Message-ID: <Pine.BSF.4.10.9903251822280.2536-100000@jig.ordway.org> In-Reply-To: <4.2.0.32.19990325105011.03a9fc30@mail.computeralt.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 25 Mar 1999, Scott I. Remick wrote: > It seems that from what you're saying, a "properly setup Unix system" is > one that simply has lots of swap space (and/or RAM) configured so that > there's never a condition that forces you to run out of swap. Yes. But what 'lots' or 'plenty' of swap space is depends on your exact situation. I see that you are concerned with comparing Unix to Windows NT. Consider that, given the same users with the same needs, most any Unix system will consider fewer resources 'plenty' than an NT system will. One nice Unix system (e.g., 256MB RAM, 10GB RAID, dual Xeon) will be able to run more services more reliably than the same machine running Windows NT. This fact is undisputed (afaik) and is true generally of Unices. So 'considering your users and their needs' also means considering your OS. > I misunderstood where the answer was. I still must say though that > throwing more RAM or HDD space at a problem in the hopes of avoiding it > seems more an answer from the Microsoft textbook and not one I'd expect > for a unix-based OS. You make a good point, but think of when MS says 'add more gear' compared to when Unix vendors say that. For NT, it's 'one box per server process' -- for Unix it's 'slap in another DIMM when hosting that 7th virtual domain'. > I don't doubt those of you who say it's difficult to program, but it's > unfortunate that we have this big hole in our team's defenses. :( There is no (relative) hole. All systems have limitations. > Oh well. I guess now after several years, I can no longer say that our > unix server has never crashed (I'm pretty much the only pro-unix > advocate in a company of NT-heads). 'After several years', indeed. :^) Amusing Unix anecdote: I administer an HP Apollo 9000 (68030) with 6MB RAM and 12MB swap. On average, ~4 users are logged in running irc(1) under screen(1) and has an X query running. Yep, it's slow (although faster than you'd think) -- but it goes down only when there's power trouble. Average uptime is ~200 days. Tell the NT-heads that. ;^) Christopher Palmer Assistant Systems Administrator, Ordway Music Theatre cpalmer@jig.ordway.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.10.9903251822280.2536-100000>