Red Hat Bugzilla – Bug 41117
Install fails on Compaq Proliant PL1600R when attempting software-based RAID1
Last modified: 2007-04-18 12:33:18 EDT
From Bugzilla Helper: User-Agent: Mozilla/4.0 (compatible; MSIE 6.0b; Windows NT 4.0) Description of problem: Installer crashes at the point where it tries to format the drives when I have a RAID1 disk pair as one of them. How reproducible: Always Steps to Reproduce: 1. Attempt to make one of the mount points a RAIDed drive 2. Continue with the install program until it is time to format the drives. 3. Kick something expensive Actual Results: The install program crashed with an error message and rebooted the computer Expected Results: It should have formatted the drives and continued with the install Additional info: Compaq Proliant PL1600R 400Mhz P2 processor 832M of RAM NCR53C8xx-based SCSI Controllers Traceback (innermost last): File "/var/tmp/anaconda-7.0.1//usr/lib/anaconda/iw/progress_gui.py", line 20, in run rc = self.todo.doInstall () File "/var/tmp/anaconda-7.0.1//usr/lib/anaconda/todo.py", line 1469, in doInstall self.fstab.makeFilesystems () File "/var/tmp/anaconda-7.0.1//usr/lib/anaconda/fstab.py", line 549, in makeFilesystems self.createRaidTab("/tmp/raidtab", "/tmp", createDevices = 1) File "/var/tmp/anaconda-7.0.1//usr/lib/anaconda/fstab.py", line 462, in createRaidTab isys.makeDevInode(device, devPrefix + '/' + device) File "/usr/lib/anaconda/isys.py", line 195, in makeDevInode return _isys.mkdevinode(name, fn) SystemError: (0, 'Error') Local variables in innermost frame: fn: /tmp/raid0p1 name: raid0p1 ToDo object: (itodo ToDo p1 (dp2 S'method' p3 (iimage CdromInstallMethod p4 (dp5 S'progressWindow' p6 <failed>
Weird. We have not seen this in our internal testing. I'll take a look.
Have you tried Red Hat Linux 7.1? I'm curious to know if you could duplicate the behavior with it. How big are the partitions that you made? Also, do you see any error messages on VC3 or VC4?
I don't have 7.1 to test, and haven't gotten around to downloading it yet. I just knocked out the Linux RAID partitions and put a normal one on one of them for now. The other one is just sitting there being ignored. That will have to change once the server goes live, but that project got put on hold pending more "critical" matters. ("Drop everything!! We broke something and the future of the company depends on you fixing it now!" *sigh* Mortals...) They are identical 4gig drives and I made the entire things one partition. (It was only meant to serve as /var/lib/pgsql/data/base.) Unfortunately, I don't remember what was on the other VC's. I assumed that the Annaconda dump would have grabbed everything I'd need.
Ok. Well, please reopen this bug if the problem reappears in 7.1
I have the exact same problem as stated above, with Redhat 7.1 Using the near exact same hardware, on a Prolaint 1600R, with Dual 450Mhz P2 processors 390M of RAM NCR53C8xx-based SCSI Controllers The installer kicks back the error similar to "a serious error has occured attempting to format /. Press any key to reboot". I tried several raid configurations: Two 9.1 gig drives partitioned several ways. If I define them as non-raid, those partitions format fine, but the installer "errors-out" as soon as it tries to format any raid partition. Thinking that it might be something to do with the drives, I pulled the two 9.1 gigs and tried the procedure on three 2.1 gigs -- same thing. It does not seem to matter if I try raid 0 or 5, the installer cannot format. Any help or comments would be greatly appreciated. sincerely yours, Hari Boukis
I also had the same problem with 7.1 and never did get it working. (I don't work for that place anymore, though, so I personally have no real interest in it anymore.)
jm3, when the crash occurs, can you press <Ctrl><Alt><F2> and type 'fdisk -l'? Please attach the results of that here.
Closing due to inactivity. Please reopen if you have more information.