Hi Lyndon! > -----Original Message----- > From: ietf <ietf-bounces@xxxxxxxx> On Behalf Of Lyndon Nerenberg > Sent: Wednesday, November 25, 2020 1:03 PM > To: ietf@xxxxxxxx > Subject: Re: Call for Community Feedback: Retiring IETF FTP Service > > John just saved me a lot of typing; let me just +1 everything he said. > > For the record, I am very much opposed to shutting down the FTP service. > > I have a cluster of systems using FTP to mirror I-D's. Just the .txt and index > files. This is trivial to perform with FTP, and I've been running the same FTP- > based mirror script for over a decade. > The cluster I use doesn't support rsync, and likely never will, yet without that > it's still more than functional enough to work as a distributed desktop > environment. I won't declare the OS here as that will almost certainly just > devolve into a set of ad hominem attacks against the operating environment, > none of which are relevant to the discussion. Understood. > I will also note that part of the reason ftp.ietf.org sees so little traffic is likelty > due to the highly aggressive session timeout on the server's FTP command > channel, making it pretty much unuseable for interactive sessions. If you > increased that timeout from 20 seconds to something reasonable, say five > minutes, people might start using the service interactively. This seems to reference another use cases -- interactive access. Is that also from the same end point identified above? If so, you've asked us not to discuss it, so ignore my ask. However, if this is a different client, can you share more details on your client. I ask because it is my understanding that most FTP clients now auto-reconnect hiding any such issues. Responding privately is fine too. As to linking this configuration to low FTP usage, I'm not so sure. As I noted in https://mailarchive.ietf.org/arch/msg/ietf/l2rQFow9G-Nc1U04dLTCzsv5ka0/, this configuration has been in place for at least a decade and no user feedback has been voiced to indicate this is a problem (until now). Regards, Roman
