Date: Wed, 12 Jul 2000 05:39:34 -0400 (EDT) From: Adam <bsdx@looksharp.net> To: Dag-Erling Smorgrav <des@flood.ping.uio.no> Cc: Satoshi - Ports Wraith - Asami <asami@FreeBSD.ORG>, ports@FreeBSD.ORG Subject: Re: fetch(1) timeout Message-ID: <Pine.BSF.4.21.0007120534530.427-100000@turtle.looksharp.net> In-Reply-To: <xzp4s5v7l0v.fsf@flood.ping.uio.no>
next in thread | previous in thread | raw e-mail | index | archive | help
On 12 Jul 2000, Dag-Erling Smorgrav wrote: >asami@freebsd.org (Satoshi - Ports Wraith - Asami) writes: >> Adding a timeout value is fine, but isn't 60 seconds a tad too short? :) > >I don't think so; the timeout value only applies to time spent waiting >for data from the server in command mode, not to total time spent >fetching the file. Receipt of any data at all from the server resets >the timeout, so if the server spends 40 seconds reacting to the PASV >command, and then 40 seconds reacting to the RETR command, libfetch >will not time out, even though the total exceeds 60 seconds. What if something like your internet connection becomes intermittant and drops for 3 minutes (ISP screwed route, cable modem/dsl unit resets, janitor trips over ethernet) and you return with the same ip? If you were 95% done downloading XFree86 at 2k/sec(bad net day?) and it timed out at 60 seconds that would be pretty dissapointing. Maybe a printf warning at 60 seconds informing of a future impending drop? (or not) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-ports" 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.4.21.0007120534530.427-100000>