This is a bit of a puzzle for me, I have a server that uses WinSCP to download files from a remote site on a daily basis, but have recently had Connection Failed messages. These seem to coincide with an upgrade taken place on the remote side, and the named host name has a new IP address and certificate behind it.
I thought this could be an issue on the remote side, however I tried connecting via my laptop instead and it worked. The difference being that I was prompted to accept the connection using the new certificate on my laptop.
I have tried setting up a new site by exporting the working one from my laptop using the IP address rather than host name, but it hasn't helped unfortunately. I believe the issue may be that it has cached the previous certificate details and is not prompting to save the details of the new one, however I am happy to listen to better ideas!
The details of the connection:
– Protocol: FTP
– Encryption: TLS/SSL Implicit
– Min TLS 1.0 , Max 1.2
Here's an extract from the log for further info:
. 2021-11-03 11:00:00.470 Connecting to XX.XXX.XXX.XX:990 ... . 2021-11-03 11:00:00.470 TLS layer changed state from unconnected to connecting . 2021-11-03 11:00:00.470 TLS layer changed state from connecting to connected . 2021-11-03 11:00:00.470 Connected with XX.XXX.XXX.XX:990, negotiating TLS connection... . 2021-11-03 11:00:15.437 Timeout detected. (control connection) . 2021-11-03 11:00:15.437 Connection failed. . 2021-11-03 11:00:15.437 Got reply 1004 to the command 1 * 2021-11-03 11:00:15.484 (EFatal) Connection failed. * 2021-11-03 11:00:15.484 Timeout detected. (control connection) * 2021-11-03 11:00:15.484 Connection failed.