Re: cvs should be up again

On Wed, Sep 15, 1999 at 02:09:09PM -0400, Mark Musone wrote: > you are not understnading what the problem is/was. > > those sockets were stuck in FIN_WAIT and FIN_CLOSE. There is no "FIN_CLOSE." The server side socket will be in TIME_WAIT or CLOSED_WAIT state, depending on which side closed the connection. Refer to RFC 793. A nice overview is also given in TCP/IP Illustrated Vol I (18.6 TCP State Transition Diagram). (requiescat in pace W. Richard Stevens) > issuing a USR1 would _not_ have fixed the problem. USR1 is a soft way to tell xinetd to try to bind again. The advantage is that other services keep actually running. > as a case in point, killing xinetd and restarting it > did not even fix the problem. Of course, you can also achieve that by restarting it. > Please dont tell me about xinetd. I helped practically > write it. Sad thing that you aren't mentioned at all in the current xinetd sources. -- Regards, Sascha Schumann Consultant