Subject: TCP gurus: explain this ssh hang...
To: None <current-users@netbsd.org>
From: Todd Whitesel <toddpw@best.com>
List: current-users
Date: 01/08/2000 08:57:16
As many of you know, I've been experimenting with ssh as a 'proper' rsh for
batch use, specifically running commands on faster machines when this can be
done safely. Tonight I was experimenting with remoting gzip (from a Sparc IPX
to an iMac) and noticed that my other session (a sun3 remoting cc1 to a cross
built cc1 on an arm32 shark) had hung for more than ten minutes, probably due
to excessive network traffic triggering something.

Here's what netstat had to say (other connections edited out, of course):

[obi-wan, the server machine, running sshd with PR-9119 fix applied]
Proto Recv-Q Send-Q  Local Address          Foreign Address        State
tcp        0      0  obi-wan.ssh            dagobah.1023           FIN_WAIT_2

[dagobah, the client machine, running ssh -c none obi-wan]
Proto Recv-Q Send-Q  Local Address          Foreign Address        State
tcp       15      0  dagobah.1023           obi-wan.22             CLOSE_WAIT

Note that my ssh's are 1.2.25, vanilla from the net and not from pkgsrc.

Todd Whitesel
toddpw @ best.com