home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!news.funet.fi!hydra!klaava!klaava!hurtta
- From: hurtta@cs.Helsinki.FI (Kari E. Hurtta)
- Newsgroups: comp.unix.programmer
- Subject: Re: Detecting a broken socket
- Message-ID: <HURTTA.92Sep7120329@plootu.Helsinki.FI>
- Date: 7 Sep 92 10:03:29 GMT
- References: <camh.715839302@balrog>
- Sender: news@klaava.Helsinki.FI (Uutis Ankka)
- Organization: Department of Computer Science, University of Helsinki, Finland
- Lines: 18
- In-Reply-To: camh@nms.otca.oz.au's message of 7 Sep 92 04: 15:02 GMT
-
- In article <camh.715839302@balrog> camh@nms.otca.oz.au (Cameron Hutchison) wrote:
- > My method of testing the keep alive timer was to have a server
- > machine "emulating" the comms server (from the standpoint of a socket
- > connection) and a client machine to connect to it. I then simulated
- > the dropping of the connection by pulling the network connection
- > to the server machine. I let the client machine run for about 45
- > minutes before giving up. The client machine did not detect the
- > broken connection.
-
- Well, If I have heard this correct: In some implementations this
- time out of keep alive timer is 2-4 HOURS.
-
- - K E H
- ( Kari.Hurtta@Helsinki.FI,
- hurtta@cc.Helsinki.FI,
- hurtta@cs.Helsinki.FI,
- HURTTA@FINUH.BITNET
- )
-