Date: Fri, 13 Jun 2003 10:59:52 -0500 From: Jamie Novak <novak@qwest.net> To: freebsd-hackers@freebsd.org Subject: "jumpstation" configuration Message-ID: <20030613155952.GA38455@qwest.net>
next in thread | raw e-mail | index | archive | help
I hope this is okay to post here. I was torn between this list and freebsd-general... At work, we have a FreeBSD jumpstart server configured, more or less, as described on Alfred Perlstein's web page. (http://people.freebsd.org/~alfred/pxe/) The jumpstation itself works great and we've been using it for well over a year to do automated installs of FreeBSD on Intel-based systems of varying types. They've all had NICs which used the "fxp" driver. Recently, however, we've started purchasing new servers that have onboard gigabit NICs which use the "em" driver. FreeBSD recognizes it and handles it just fine (tested using a CD install), but we're running into a problem with the network boot/install that I thought maybe somebody else could help me with. On the fxp servers, they'll network boot and install without complaint, so we know we have DHCP, tftp, nfs, etc. working (I would assume) the way they should. Our the em-based servers, the jumpstart process will get an IP address from DHCP, it'll get the kernel image from tftp, it'll go through the boot screen and get to the actual FreeBSD install.cfg (sets a bunch of variables, the hostname, the netDev=em0, etc.) but as soon as it gets to "mediaSetNFS" in the install.cfg, it dies without warning or error -- the install screen just says "Command `mediSetNFS' failed." Does anybody know what sorts of things I could look for to figure out why it gets this far, but then fails? (As I mentioned before, nfs appears to be configured correctly, because the fxp-based servers blow right past that point and get to the actual install without a problem.) I don't even know where to start... - Jamie
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030613155952.GA38455>