The fun never ends. TO recap, I installed SLES 8 onto a dual Opteron 244 system running on a MSI K8D Master board. I was getting hangs on bootup when the process gets to setting up the IDE disks. I found that I was able to boot into safe mode every time, but a full boot was nearly impossible. I suspected the Adaptec 29160 HBA was interfering with it, so when I removed it, the system boots up. So I brought the system up this way, and our DBA was able to install Oracle 9, and create a test database. During one of the times that the system booted up with the 29160, I found that it had trouble working with either the disk or the file system on the disk that was more than 1TB in size. So far, this OS is not impressing me at all. I also saw a whole bunch of emails about upgrading to Service Pack 3, so that's what I did. Lo and behold, the boot problems went away, and the system seemed to be able to work with disks greater than 1 TB. All seemed fine. The only problem is that the SCSI HBA is not automatically configured into the system, and I have to force load the aic-7xxx driver from a script in /etc/init.d/bootd. But now, as my DBA tries to install another database, a real one, he finds the Java based install procedure always crashing at the same place. The message that pops up shows an error while trying to write to some Java library file. The file appears to be there, but since the procedure is so opaque, we have no idea what it's trying to do. We tried different versions of JDK, with the same results. The tech support at Oracle says that these types of error messages in association with SP3 on SLES 8 usually indicates that not all of the kernel patches have been installed. Yet when I use YOU to check the SuSE site for updates, it would find no more patches. I have just re-installed the SLES 8 off of my CDs, and it is refusing to boot up. It is consistent with my suspicion that when the install procedure configures the 29160 card (or any other SCSI card) into the initrd, the system simply will not boot. The console message is: Unable to handle kernel NULL pointer dereference The machine does not respond to any kind of interrupts in this state, and I have to hold down the power button to shut it off. But oddly, if I select Safe Boot from the GRUB menu, it boots up, but in obvisouly limited mode. What is the difference? So the situation is, if I don't apply these patches, I can not boot. If I do apply these patches, I can not install our primary application, Oracle 9i. If I don't find a solution soon, I will have to ask SuSE for my money back on this wonderful product, and try the Red Hat Enterprise Server. My colleagues are now very wary of the situation as well, which sort of reinforces their original preference for RHES. (Yes, they're saying "I told you so") Any ideas? eyc