Date: Sun, 31 May 1998 19:36:57 -0700 From: Mike Smith <mike@smith.net.au> To: "Andrew Specht" <andrew@iaccess.com.au> Cc: freebsd-current@FreeBSD.ORG Subject: Re: mbuf cluster problem continues!! Message-ID: <199806010236.TAA14834@antipodes.cdrom.com> In-Reply-To: Your message of "Mon, 01 Jun 1998 10:28:03 %2B1000." <015b01bd8cf4$23f4da40$e34a05cb@alpine.iaccess>
next in thread | previous in thread | raw e-mail | index | archive | help
> Hi again, > > I'm still having the same mbuf cluster problem. I'm running squid with a 14 > Gig Cache and getting up to 200 connections a second. The problem is that > mbuf clusters in-use just keeps on rising until it gets to the peak and then > the whole thing crashes. I've got it set to 22000 at the moment, but last > time they went up to nearly 10000 before it crashed. Is there a problem > with leaking mbuf clusters still, or is that what they are supposed to do? Are you actually running on -current? And this is a production system? Are you aware that this is a really bad idea? The issue with mbufs is that you can't dynamically allocate them when you need them, so you have to have enough to begin with. Unfortunately, when you reach the point where you need them and there aren't any left, you can't (often) fail gracefully - things expect that they are going to get them when they need them. This is a fairly major flaw in the system's design. 8( However, I am not aware of any known mbuf leaks in the system at the moment. Are you certain that they're being leaked? Has your load managed to bring the system down with the settings at 22000? -- \\ Sometimes you're ahead, \\ Mike Smith \\ sometimes you're behind. \\ mike@smith.net.au \\ The race is long, and in the \\ msmith@freebsd.org \\ end it's only with yourself. \\ msmith@cdrom.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199806010236.TAA14834>