After updating my test3 rawhide to today updates, and after reading about kexec, I decided to try it: impressive! My system is a dual amd mp with serial ata pci card, scsi pci card and other cards/devices, so bios startup phase is quite annoying, especially at 01:39 AM... ;-) And also my main bootloader is grub on CentOS 4 on first disk, with option to chainload grub of FC5, installed in its root partition on second hd... so two grub splash screens are too many at this time... So my runnning kernel was 1955smp and just installed updates and in particular the 1996smp kernel. I ran: 1) kexec -l /boot/vmlinuz-2.6.15-1.1996_FC5smp --initrd=/boot/initrd-2.6.15-1.1996_FC5smp.img --append="ro root=/dev/sdb2 rhgb quiet" to load new kernel into memory 2) shutdown 0 to put system in single user mode and stop various services 3) kexec -e to reboot to the new loaded kernel All has gone well with immediate boot up of the kernel! Only one doubt: at boot obviously the root filesystem appears uncleanly closed and so I get the messages: EXT3-fs: INFO: recovery required on readonly filesystem. EXT3-fs: write access will be enabled during recovery. kjournald starting. Commit interval 5 seconds EXT3-fs: sdb2: orphan cleanup on readonly fs ext3_orphan_cleanup: deleting unreferenced inode 1263797 ext3_orphan_cleanup: deleting unreferenced inode 819283 ext3_orphan_cleanup: deleting unreferenced inode 819224 EXT3-fs: sdb2: 3 orphan inodes deleted EXT3-fs: recovery complete. EXT3-fs: mounted filesystem with ordered data mode. Of course, as I was in init 1, "bowls were quite slow", but the orphan clean up was not the marvellous message I like... DO I have means to avoid this situation? Furthermore with FC5 you don't get the option to check filesystem on unclean shutdown and so only replay of journal is executed (correct?) Thanks for your time Gianluca -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
