RE: [suse-security] SuSE Security Announcement: kernel (SuSE-SA:2003:021)
Do you have local access to the system or a serial console? If yes, then just try it. If the kernel failed you can immediatelly go back to the old kernel. So you at least don't have a regression.
Hi there could you explain the 'immediatelly go back' a little? My Server is a Suse 7.2 to which I don't have local access, but i can boot my server remote with a rescue-debian installation, that can access the Harddisc. Do you think that would be usefull when reboot fails? if so : which files do i need to backup, so that i can restore the old kernel without a running suse on the box? Any help appreciated, since i'm not used to update kernels on remote servers and i'm a little bit in fear that the reboot will fail. thanks Sebastian
On Wed, Mar 26, 2003 at 07:29:10PM +0100, Sebastian Stadtlich wrote:
Do you have local access to the system or a serial console? If yes, then just try it. If the kernel failed you can immediatelly go back to the old kernel. So you at least don't have a regression.
Hi there
could you explain the 'immediatelly go back' a little?
Sure.
My Server is a Suse 7.2 to which I don't have local access, but i can boot my server remote with a rescue-debian installation, that can access the Harddisc. Do you think that would be usefull when reboot fails?
Yes. You can also run rpm on debian. But it is much easier, if you make a backup of /boot/vmlinuz, /boot/initrd and /lib/modules/2.4.19-4GB before. Restoring those three items and reinstalling lilo will make your system booting the old kernel again. After this reboot with the old kernel to the SuSE system you should reinstall the old kernel package, to do some cleanups. (reinstalling old System.map and so on)
if so : which files do i need to backup, so that i can restore the old kernel without a running suse on the box?
See above. Robert -- Robert Schiele Tel.: +49-621-181-2517 Dipl.-Wirtsch.informatiker mailto:rschiele@uni-mannheim.de
-----BEGIN PGP SIGNED MESSAGE----- Hi Robert and List, one question again. In the case not having done a backup and want to ga back to the unpatched kernel, is it sufficient to uninstall the patch via RPM and do the mk_initrd && lilo? Just to be sure... Thanks for an answer, although it is not a genuine security question: Tobias Am Mittwoch, 26. März 2003 19:44 schrieb Robert Schiele:
Yes. You can also run rpm on debian. But it is much easier, if you make a backup of /boot/vmlinuz, /boot/initrd and
- -- - ------------------------------------- Tobias Cremer Hohenzollernstr. 20 80801 München to.c@gmx.net http://www.tobiwan-kenobi.de -----BEGIN PGP SIGNATURE----- Version: 2.6.3in Charset: noconv iQCVAwUBPohiDrd9JiPf/I1hAQG7eQP+PlMNSDuX73brNnpCERHnS0/D/VikNtWC uU4C+1YRmPWdkHCZkHkano1xZm17dksZ1UNapjPiRyS1yVoIxPwPoOZKqNNOOfPI PK5U/+07oGJiwYG2jL4R5k1YgS7zEK0wXznIgRwkWm64iIyIWepygfiPtBpuCH9d R/2LJBUSvGU= =DhHy -----END PGP SIGNATURE-----
CAn anybody says me where is the pach for this problem? thanks
IMHO: SuSE now has own fate... Too many release makes difficult to publish all security patches quickly, and rapidly come some release into "discontinued" state (as my 7.0 ...). One release per year, but a "perfect" one makes me happy. ;-) Ferenc ------------------------
participants (5)
-
Dionisio Ruiz de Zarate -
Robert Schiele -
Sebastian Stadtlich -
tobiwan -
Ősz Ferenc