From owner-freebsd-stable Mon Feb 5 9:41:29 2001 Delivered-To: freebsd-stable@freebsd.org Received: from fw.wintelcom.net (ns1.wintelcom.net [209.1.153.20]) by hub.freebsd.org (Postfix) with ESMTP id C362237B401 for ; Mon, 5 Feb 2001 09:41:11 -0800 (PST) Received: (from bright@localhost) by fw.wintelcom.net (8.10.0/8.10.0) id f15HfAZ03616; Mon, 5 Feb 2001 09:41:10 -0800 (PST) Date: Mon, 5 Feb 2001 09:41:10 -0800 From: Alfred Perlstein To: Marius Cc: stable@FreeBSD.ORG Subject: Re: Multiple dropped thttpd connections under high load. Message-ID: <20010205094110.O26076@fw.wintelcom.net> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: ; from marius@mail.communityconnect.com on Mon, Feb 05, 2001 at 11:32:20AM -0500 Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG * Marius [010205 08:33] wrote: > > > To anyone who has been following my little trials with thttpd 2.17 and > FreeBSD 4.x stable, I have one more piece of news. After remaking world, > and recompiling our customized thttpd from scratch, we are still not > working. In fact, I think it is worse. A ktrace and a truss both > indicate that thttpd not surving very long, 10 seconds or less, then > the process completely dies. The perl wrapper around thttpd had just been > restarting it, so we didn't really notice at first. Obviously that is the > source of our dropped connections when the load gets high enough. > > Since other people are uning thttpd on the 4.x branch, I doubt thttpd > itself is to blaime. Perhaps some of the patches to thttpd we added > arn't playing well with 4.x, or perhaps somthing has gone wrong on a > hardware level on that machine. I might back ot to 3.5-S and see if it > exhibits the same behavior. Either way, more testing/checking is called > for on my part. > > I do thank the list for the input they have given me. I certainly don't > want to appear as ungrateful. We had the same problem as well, my guess is that thttpd is exiting due to a weird errno result returned by one of the socket functions. If the thing only lasts 10 seconds before dying, you should be able to run truss/ktrace on it and figure out where it's dying. We had the same problem, then I wrote my own image server and ditched thttpd, if your company has a small amount of content and wants to license my server you can reply in private mail. It's doing ~260 connections a second and using approx 1.5% cpu. -Alfred To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message