I know 425 is the right thing to do here, but FileZilla gets stuck on it and I need a way to recover from this error.
Thanks Sachin -----Original Message----- From: Niklas Gustavsson [mailto:nik...@protocol7.com] Sent: Sunday, September 23, 2012 9:13 PM To: ftpserver-users@mina.apache.org Cc: Sachin Shetty Subject: Re: GnuTLS internal error in Filezilla On Sun, Sep 23, 2012 at 4:42 PM, Sachin Shetty <sshe...@egnyte.com> wrote: > patched STOR.java to catch the SSL Exception and send a 426 instead and > throw the exception. This tricked filezilla in to retrying instead of > getting stuck. Since this happens on creating the socket rather than while transferring data, sending a 425 response is probably more appropriate as is already done on Exception. But I guess that will mean that Filezilla will get stuck? Any reason why you retrow the exception? I gather that will cause multiple responses to the client? > 4. Should we consider upgrading to the latest 1.1.0-SNAPSHOT - how risky > would that be? In general, 1.1.0-SNAPSHOT is not stable and thus I would refrain from running it in production. So, I would recommend to attempt to backport your custom code to the 1.0.x branch if at all possible. That said, I've browsed the diff between 1.0.x and trunk (1.1.0-SNAPSHOT) without finding what would cause this. > 5. Where the heck is this 1 hour thing coming from - I have made sure > firewall etc is out of the picture and still cannot explain it. I would guess it something with the SSL handshake that fails and times out after 1 hour. /niklas
