On Sun, May 05, 2013 at 08:26:11PM -0400, Felix Miata wrote:
On 2013-05-05 16:30 (GMT-0700) Lew Wolfgang composed:
Wow, that was nasty! I pulled a zypper dup a few hours ago and it completely borked my 12.3 system. Zipper failed with three kernel debug models that ran out of space in /boot. Since I had issues with reiserfs some years ago I've been configuring /boot as a 100-MB ext2 filesystem. Alas, 100-MB is no longer big enough!
When I first started multibooting with Linux I was going by RedHat's minimum supported /boot size of about 70M. Not terribly long after that I shifted to 200M. A year or so ago I upped again, to 400M.
Of course, each /boot on mine are just a directory in each /. It's the real boot partition, which I never mount to /boot, that I make 400M. It provides a home for a master Grub and a number of installation kernels and initrds for doing HTTP installations. I only put regular operating kernels & initrds there on RAID systems, by copying them there from where the various package systems put them.
As to usage of /boot, the 13.1 I currently have booted has two installed kernels, 5 initrds, contains none of Grub2 (Legacy only), and consumes 57,343 1024kb blocks, or about 55MB.
FWIW, in openSUSE 12.3 we have enabled "multiversion" kernels, so old kernels will remain on update. It is currently set to: multiversion.kernels = latest,latest-1,running The clean up is done by the purge-kernels init script in "mkinitrd". As we noticed only last week, the "clean up" job is not running, so you should run once: systemctl enable purge-kernels Ciao, Marcus -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org