Created attachment 817173 [details] storage.log Description of problem: The installation media is a F19 x86-64 netinst CD-R that passed boot media test. Anaconda's GUI shows "No disks detected", while two hard drives are shown in dmesg as sda and sdb. SATA is configured as AHCI in BIOS. Machine is a Colfax blade server. HDs are two ST9500620NS. Version-Release number of selected component (if applicable): Official F19 netinst x86-64 CD. How reproducible: Always Steps to Reproduce: 1. Boot machine by using the F19 netinst x86-64 CD 2. Select US keyboard 3. Go to select disk and see "No disks detected." 4. Press Ctrl-Alt-F2 to switch to shell and run dmesg, see two hard disks are properly detected as sda and sdb. Actual results: No disks detected by Anaconda. Expected results: Two hard drives should be detected. Additional info: In storage.log, sda is detected as "part of a biosraid". I have no idea why that is. In BIOS the SATA disks are set as AHCI, not RAID. The machine worked well running F18 before. This bug might be a duplicate of #947100.
Created attachment 817174 [details] syslog
No disks are selected because you didn't select any disks. Based on the storage.log, there should be two disks available when you enter the installation destination spoke. If this not the case, reopen this bug.
David, read carefully. The OP said no drives were DETECTED, not selected. If he is having the same issue I am, it's that when you go to select the drive to install to, it isn't there, even though it is detected elsewhere, such as dmesg and /dev.
(In reply to David Shea from comment #2) > No disks are selected because you didn't select any disks. Based on the > storage.log, there should be two disks available when you enter the > installation destination spoke. If this not the case, reopen this bug. That is not the case. Anaconda didn't display any disk for me to choose.
After some additional research and experimenting, I have a solution, which is not to say a bug does or doesn't exist, but for my purposes, its acceptable. First, some additional context: the Anaconda installer was only showing one of three hard drives attached to the system. All three are MBR drives, as opposed to GPT. Output from' fdisk -l' showed that the first sector of the two drives that were not appearing was 63, and for the one that was, 2048. After moving the first partition a couple MB from of the beginning of the partition table on an affected drive, it now appears in Anaconda. The issue appears to be directly related to where the partition table starts, but as far as I'm concerned, burning a meg or two is trivial. The biggest inconvenience is waiting on the partition to resize. Also, I would expect this problem to become less common as more drives are switching to GPT and not being originally formatted under Windows XP, as was my case. Anyway, hope this helps.
The storage log attached shows sda1 and sdb1 starting at 2048 sectors, so I that is not the issue. Does this problem still occur with F20 or rawhide?
We have been using dnf upgrade since then and haven't installed any new system yet so I haven't gotten a chance to test. Closing this for now.