Red Hat Bugzilla – Bug 481202
Experiencing large number of errors when performing RHEL 5.3 Upgrade
Last modified: 2010-11-09 08:09:58 EST
Created attachment 329727 [details] Finally seeing a normal screen warning message after 20 minutes of errors. Description of problem: I am experiencing something different with RHEL 5.3. This is on a IBM BladeCenter with 12 blade modules. Eight of the blade modules are part of 8-node RHEL 5.2 Cluster and four of the blade modules are part of a 4-node RHEL 5.3 Test cluster. I was trying to do an upgrade of RHEL 5.3 Beta to RHEL 5.3 using the DVD media. The install boot is hanging on various drivers : - pdc - asr - ddf11 and drilling down through all shared scsi disks with errors similar to this: ERROR: pdc: reading /dev/sdNN [Input/Ouput Error] Version-Release number of selected component (if applicable): How can I skip this repetitive scan of the shared disks? How reproducible: Simply attempt to install Product. Steps to Reproduce: 1. Boot off of DVD MEdia or boot.iso to do a network boot. 2. 3. Actual results: It took 20 minutes to get to the "Warning" prompts about initializing all data on various shared disks which of course I do not want touch, i.e., "NO". Expected results: Should only take a minute to get to this point. Additional info: Image attached.
Created attachment 329729 [details] Searching for .... Installations
It should be noted that IBM's RDAC driver was installed in order to access an IBM DS4700 Disk Shelf while running RHEL 5.3 beta. mppUtil (1) - RDAC (a.k.a. MPP) Driver Utility sg_rdac (8) - Display or Modify RDAC Redundant Controller Page ime3/root> ls -ld /usr/local/src/ibm.rdac/linuxrdac-09.03.0C05.0030 drwxr-xr-x 5 bin bin 4096 Nov 12 15:59 /usr/local/src/ibm.rdac/linuxrdac-09.03.0C05.0030
The IBM RDAC instructions say you should uninstall the RDAC driver from current kernel and then boot off of new kernel and reinstall RDAC driver. This did not work as well. The odd thing, I was unable to boot off of the RHEL5.2 DVD, RHEL 5.3 DVD, RHEL 5.2 boot.iso FTP, RHEL 5.3 boot.iso, etc. So, I gave up and finally offlined all of the blade modules, offlined the DS4700 shared disks, before I could boot onto the RHEL 5.3 RC kernel. I have no idea why this is happening. Before, the install cycle would simply prompt you and ask did you want to initialize each disk and continue. It would be nice if there was a way to restrict the probe of the shared disks, regardless.
Hi Everett, You've filed this bug against the Red Hat Enterprise Linux Installation Guide -- in other words, as if it were a mistake in the documentation of the installation process rather than a problem with the process itself. I'm changing the component to "anaconda", the Red Hat Enterprise Linux installer. Ruediger
That's okay. However, I am now trying to recover a IBM blade by booting of RHEL 5.3 x86_64 media via Memory Management software.....it's hanging on the shared disks.. Any comments or suggests?
Created attachment 337397 [details] Trying to use rhel53 x86_64 dvd in rescue mode, i.e. linux rescue
Has anyone had a chance to review? Requesting an update.
Hi Everett, you should work this through your normal support channel if it is a high priority. They can provide troubleshooting assistance.
Upgrading from a beta release is not supported.