From mboxrd@z Thu Jan 1 00:00:00 1970 From: ron minnich To: 9fans@cse.psu.edu Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Subject: [9fans] silly question on ssh Date: Tue, 15 Apr 2003 08:35:29 -0600 Topicbox-Message-UUID: 900649a6-eacb-11e9-9e20-41e7f4b1d025 I have a linux box running several vmware images of plan 9. One plan 9 image, running a term and login as glenda, can ssh to the linux side no trouble. The other, a cpu/auth/kfs server with me logged in as user bootes, used to work, but nowadays is getting 'client authentication failed'. tcpdump shows the client side (plan 9) closing the connection. What's a good way to start debugging this kind of thing? My keyring is empty. I'm not really sure where to start, and am ssh-impaired to boot. Another silly question. Same box. I have a cpu node (real hardware) coming up with 9load in flash. It is connected to eth0, 10.0.4.1. The vmware network is vmnet8 @ 172.16.189.1 (vmnet8 as vmware says you have to use NAT if you have wireless, which I do). I am also routing-impaired :-) I have tested this cpu hardware downloading plan9 kernels via 9load and a dhcp/tftp server in linux, and it works. Now I need to have it connect to the auth server in vmware, and I would most prefer to have it suck the kernel down from the auth server in vmware. What I want is to rig things somehow so that the DHCP requests from the hardware CPU node on the eth0 net (10.0.x.x) go to the vmware auth server on vmnet8 (172.16.189.x). I note that broadcasts are not making it from the 10.0.x.x net to the 172.16.189.x net. OK, anyone know the magic formula to make this work? How many things am I doing wrong-- countably or uncountably infinite? ron