>3. I had this weird problem, when I setup the cluster, all clients network >booted and pulled image off the head node fine, and then booted off the >hard disk fine, then I made some changes (add and configured NIS and >autofs, disable the iptables/pfilter), then rebooted, this time the >network was gone, it couldn't find eth0 interface, ifconfig only showed >loopback interface, then I found out there was no /etc/module.conf file, >so the 3c59x module was not loaded, is this strange, or just me?
This can happen if your eth0 driver isn't included in the Oscar distribution and you had to do a specialized installation. The specific line you want in the module.conf file is 'alias eth0 3c59x'. I think there are two options: (1) login to the console on each node, create module.conf (for future reboots), and then do 'insmod 3c59x' to bring up the network now. If that doesn't work, you may have to restart the network service; or (2) create module.conf in the oscar image override directory, and reinstall oscar, starting with the client detection step. >4. On all of my clients, the "xfs" process taking 100% CPU time, anyone >has any idea? I am looking into it now. We noticed that also. You can disable it. Jay ------------------------------------------------------- The SF.Net email is sponsored by: Beat the post-holiday blues Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek. It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt _______________________________________________ Oscar-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/oscar-users
