eho@reason.princeton.edu (Eric Ho) (03/30/89)
Does anyone has any problems with 4.0 ifconfig under a 3.3 environment ? I've recently tried to upgrade my Sun-4 from SYS4-3.2 to 4.0. My Sun-4 is configured as a standalone machine and as a YP client. The rest of the net is a bunch of Sun-3's all running 3.3 and there is only 1 YP server on the net. Princeton has a Class B network and is subnetted. All the machines in our lab live under 128.112.162 and the IP address of my Sun-4 is 128.112.162.14 When I configured my Sun-4, I commented out the "ifconfig ..... +" in the rc.local file since my 3.3 YP server doesn't support that. Instead, I went into the rc.boot file and changed the ifconfig line to :- ifconfig ie0 $hostname -trailers netmask 255.255.255.0 up All other Sun-3's on my local net have the same ifconfig line in their rc.boot files. Now, the strange thing is that after my Sun-4 is up and if I then reboot my Sun-3 diskless node, it got the boot programs ok and it got vmunix ok and then it fails to do nd boot. Got the "nd: output error 51" instead. The strange thing is that after vmunix checked all the devices, it would say "Setting subnet mask to 0xffff0000" and then tried to do nd boot (i.e. boot up /dev/nd0 and /dev/ndp1). This is not what it used to be. It used to say "Setting subnet mask to 0xffffff00" instead. Then when I powered down my Sun-4 and tried reboot my Sun-3 diskless again. It booted up without a hitch and it said "Setting subnet mask to 0xffffff00" during booting like it used to do before. It seems that if my Sun-4 was up while my Sun-3 diskless was booting then my Sun-3 diskless would talk to my Sun-4 and indeed I found that out using etherfind on some other Sun-3 nodes during this booting period. Does anyone has any pointers as to why this is going on or has similiar problems ? What I really want to know is if this is a problem with mixing 4.0 with 3.X or somewhere else entirely different or that I did something wrong with my ifconfig lines. I've now backed out my Sun-4 to SYS4-3.2 and everything is back to normal again. Eric Ho Princeton Cognitive Science Lab., Princeton University email = eho@phoenix.princeton.edu voice = 609-987-2987 -eric-