Jim Being "reliable" includes giving up - and saying so - when "slings and arrows" intervene.
Perhaps I picked up the wrong emphasis from your original post. I took it to be a concern over FTP rather than a concern over the Internet. As you have demonstrated to yourself, the Internet - as they say these days - has "reliability issues"! At some point, TCP will give up retrying to compensate for the unreliability of the underlying TCP connection path. Chris Mason On Fri, 11 Sep 2009 10:26:20 +0100, Jim McAlpine <[email protected]> wrote: >On Fri, Sep 11, 2009 at 9:56 AM, Chris Mason <[email protected]>wrote: > >> Jim >> >> FTP uses TCP so it's the reliability of TCP that matters here. Of what >> underlying facilities happen to consist is very probably of no relevance. >> Thus, >> assuming the underlying facilities are inherently *less* reliable that the >> reliability offered by TCP, you get the same level of reliability whether >> you are >> passing data within your local infranet or over the Internet. >> >> So, if you consider TCP offers sufficient reliability, you don't need to go >> looking for another technique for "moving" your dumps. I suspect that, for >> dumps, TCP is quite adequate. >> >> Chris Mason >> > >Chris, thanks, that's good to hear. However I've just tried to FTP my first >dump which was approx 1.4 GB and it failed about one third the way through >with - > >Netout : Connection reset by peer >451 Transfer aborted due to receive error > >which doesn't bode well. > >Jim McAlpine ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
