Status > OpenVPN hangs when using TCP w/OpenVPN Int. "any"
When using TCP with OpenVPN, even where specifying 'local' to bind to a specific IP, the server binds itself to 127.0.0.1 as well. Then the Status > OpenVPN page hangs because the OpenVPN server is binding to the management port on 127.0.0.1. So the status page actually connects to the OpenVPN server rather than the management interface, which returns nothing to its queries.
Three issues here.
1) If the OpenVPN management interface does not respond, it should give up after several seconds, rather than completely hang up the web interface.
2) When using "any" with TCP, the management interface must use a different port than the one OpenVPN is using.
3) why is the server apparently binding to 127.0.0.1 where it shouldn't be (or is that not actually the issue)
Ticket #826. Add timeout of 1 second for all read/write actions performed on the socket. This should fix point 1) on the ticket.
Ticket #826. Convert to unix domain sockets for management interface so we do not have problems when interface is any.
#1 Updated by Jim Pingle about 9 years ago
The binding seems to be correct, but the status connection wasn't being parsed properly.
The status command to the management daemon doesn't work correctly for static key servers, you get back this error:
ERROR: The 'status' command is not supported by the current daemon mode