Description of problem: My understanding is that the rescue CD has the anaconda stage2 installer on it. However, anaconda is ftping the stage2 image from the network install ftp server given in the installer. Version-Release number of selected component (if applicable): rawhide 2007-04-17 08:31 rescue cd ftp install How reproducible: yes. also on a hp compaq nx6320 {1GB ram 1.8GHz, 60G wanting to upgrade. Steps to Reproduce: 1. boot rescue CD 2. choose network install 3. enter ftp, and path. Actual results: vt3 Transferring ftp:.../image/stage2.img to a fd mntloop loop0 on /mnt/runtime as /tmp/ramfs/stage2.img fd is 24 LOOP_SET_FD failed: Device or resource busy Error mounting /dev/loop0 on /mnt/runtime (Device or resource busy). Expected results: Since the stage2 is on the rescue CD, it should have mounted the CD to get the stage2 installer. Additional info: In the case where the stage2 image isnt present locally, the installer failed to mount the downloaded stage2.img, and after displaying error dialog, returns to the FTP Setup screen. PentiumII 366MHz, 384 MB ram, 20GB disk. Doesn't get to vt2 being a usable console.
Created attachment 152801 [details] hp omnibook 4150 lspci
Created attachment 152802 [details] hp omnibook 4150 dmesg
Created attachment 152803 [details] hp omnibook 4150 lspcmcia
Created attachment 152804 [details] grub.conf / fstab / grub for info
actually, could be similar to bug 229704, but showing up differently due to the CD boot method ?
Downloaded 2007-04-18 08:25 efdf1ea5f2a5d97be42fc977910384505e833dfa FC-development-i386-rescuecd.iso sha1sum is OK. k3b verifies CD OK. Confirm that same is occurring. Tries to mount cdrom:/images/stage2.img, but fails: {visible messages - no vt2 shell yet - can you get one ?} 23:17:12 INFO: trying to mount CD device scd0 23:17:16 INFO: mntloop loop0 on /mnt/runtime as /mnt/source/images/stage2.img fd is 20 23:17:17 DEBUG: MD5SUM -> a258a71b15afdd8a155d24d575cb3441 23:17:17 DEBUG: skipsectors = 15 23:17:17 INFO: supported = 1 23:17:17 INFO: transferring ftp://download.fedora.redhat.com/pub/fedora/linux/core/development/i386/os/.discinfo to a fd --meanwhile on vt4 <7>ISO 9660 Extensions: Microsoft Joliet Level 3 <4>Unable to load NLS charset utf8 <4>Unable to load NLS charset utf8 <7>ISO 9660 Extensions: RRIP_1991A <4>Unable to identify CD-RPM format. <4>VFS: Can't find an ext2 filesystem on dev loop0. ==== on VT1 click OK to the server/path dialog again. vt3: 23:23:01 INFO: trying to mount CD device scd0 23:23:01 INFO: transferring ftp://download.fedora.redhat.com/pub/fedora /linux/core/development/i386/os/images/updates.img to a fd 23:23:03 INFO: transferring ftp://download.fedora.redhat.com/pub/fedora /linux/core/development/i386/os/disc1/images/updates.img to a fd 23:23:05 INFO: transferring ftp://download.fedora.redhat.com/pub/fedora /linux/core/development/i386/os/images/updates.img to a fd 23:23:07 INFO: transferring ftp://download.fedora.redhat.com/pub/fedora /linux/core/development/i386/os/disc1/images/product.img to a fd 23:23:09 INFO: 381092 kB are available 23:23:09 INFO: transferring ftp://download.fedora.redhat.com/pub/fedora /linux/core/development/i386/os/images/stage2.img to a fd --meanwhile vt1: Retrieving images/stage2.img for some time and then: Unable to retrieve the install image. OK. on vt3: 23:28:49 DEBUG: copied 81604608 bytes to /tmp/ramfs/stage2.img (complete) 23:28:49 ERROR: LOOP_SET_FD failed: Device or resource busy 23:28:49 ERROR: Error mounting /dev/loop0 on /mnt/runtime (Device or resource busy) vt1: back to server/path dialog. ~~~~~ My interpretation: - the rescue CD does have the stage2.img, but is unable to mount it. - instead of erroring appropriately {something wrong with mounting}, it moves to try to get the stage2.img from the ftp server. - the ftp download succeeds. (complete) - it attempts to mount the dl'ed stage2, and fails again (same reason I guess) - user is then in an endless {time consuming and download wasting} loop.
This is repeatable on a totally different machine: desktop, amd athlon 2600, ide only, 3xdisks, with same rescue as comment #6. $ cat /etc/fstab /dev/VolGroup00/LogVolslash / ext3 defaults 1 1 LABEL=/boot1 /boot ext3 defaults 1 2 devpts /dev/pts devpts gid=5,mode=620 0 0 tmpfs /dev/shm tmpfs defaults 0 0 /dev/VolGroup00/LogVolhome /home ext3 defaults 1 2 ... ie lvm for the / partition - I don't know if that can be labelled like /boot can ? On both machines, the disk(s) is detected, during the rescue boot up.
Repeated with: $ ls -l FC-development-i386-rescuecd.iso -rw-r--r-- 1 root root 103387136 Apr 19 14:21 FC-development-i386-rescuecd.iso $ sha1sum -c SHA1SUM FC-development-i386-rescuecd.iso: OK # sha1sum F* 59747fcdeba04c1db6d831aa37810e3ee00b8287 FC-development-i386-rescuecd.iso Now the stage2 installer _is_ loaded from cdrom {ie Local Installation Media detected}. Noted also that rescue ISO grew more than 10MB today. Trying to use the upgrade option, the ide hard disk is detected {/dev/sda3}, and it looks as if install would be OK (except I have lvms without labels}.
rescue cd boots and gui loads stage2 from CD also on an Dell Poweredge 1600SC with scsi drives. Looks good.
552905763371fee09fb6bdfa5b1f2015693b93f6 FC-development-i386-rescuecd.iso from 2007-05-02 has no problem on this same machine {hp omnibook 4150 notebook}