If there is I/O to drive(s) on a controller and fdisk is run on ANOTHER controller, the fdisk will hang until all I/O has completed or been killed. The fdisk will then complete, but the keyboard may not respond afterward. Example: Configuration: QLogic 1280 w/2 9.1GB SCSI disks: sda & sdb Adaptec (Poseidon) w/2 9.1GB SCSI disks: sdc & sdd # dd if=/dev/zero of=/dev/sda bs=64k & # dd if=/dev/zero of=/dev/sdb bs=64k & # fdisk /dev/sdc Create a new partition on sdc. After selecting "w" to write the new partition table fdisk will hang after displaying the message: Calling ioctl() to re-read the partition table. ps aux shows fdisk blocked waiting on I/O. It will hang there until the dd's on the other controller complete or are killed. After killing the dd's the keyboard may not respond for several minutes, or may not come back at all. After using ssh to remotely login, the following lines were noted in dmesg: keyboard: Timeout - AT keyboard not present? keyboard: Timeout - AT keyboard not present? keyboard: Timeout - AT keyboard not present? keyboard: Timeout - AT keyboard not present? keyboard: Timeout - AT keyboard not present? The system required a reboot to recover. When trying the same test using parted, parted complained that it did not recognize the disk label "/dev/sdc". The keyboard also quit responding during this test. ia32 systems do NOT exhibit this behavior.
This defect is considered MUST-FIX for Fairfax.
What's your root partition for this?
Also, do you have quotas enabled?
My root partition was on an array controller. No quotas have been set.
Thanks for the bug report. However, Red Hat no longer maintains this version of the product. Please upgrade to the latest version and open a new bug if the problem persists. The Fedora Legacy project (http://fedoralegacy.org/) maintains some older releases, and if you believe this bug is interesting to them, please report the problem in the bug tracker at: http://bugzilla.fedora.us/