RAID Problem

Discuss usability issues, general maintenance, and general support issues for a grsecurity-enabled system.

RAID Problem

Postby nikj » Fri Jun 27, 2003 3:53 am

Hi

Apologies in advance for incomplete information.

I've been playing with a secure linux distro (http://www.openna.com) and have come across a problem with, I think, the AMI Megaraid controller and grsecurity.

The system is a Dell Poweredge 2400 with the perc/2 (AMI) RAID controller. It has 3 SCSI disks in a RAID 5 array.

Whilst I can install this distro without problems, I cannot boot it, and get the following message:

grsec: attempted resource overstep by requesting 4096 for RLIMIT_CORE
against limit 0 by (linuxrc:8) UID (0) EUID (0), parent (sqpeer:1) UID (0)
EUID (0)
kmod: faild to exec /sbin/modprobe -s -k block-major-72, error = 2
VFS: Cannot open root device "/dev/sda5" or 08:05
please apped a correct "root=" boot option
kernel panic: VFS: unable to mount root fs on 08:05

Comparing the boot sequence of this distro with the boot sequence of Redhat 9 on the same hardware, I am 99% certain that the problem here is with the AMI Megaraid module.

The kernel is 2.4.20, but I'm afraid I do not know the version of grsec as the distro maintainer is on holiday.

I am a complete newbie with grsecurity, and I don't really understand the message beyond the suggestion of a core dump. Is grsec merely reporting a problem here, or causing one?

Any ideas?

Many thanks

Nik
nikj
 
Posts: 2
Joined: Fri Jun 27, 2003 3:36 am

Re: RAID Problem

Postby PaX Team » Fri Jun 27, 2003 5:09 am

nikj wrote:grsec: attempted resource overstep by requesting 4096 for RLIMIT_CORE
against limit 0 by (linuxrc:8) UID (0) EUID (0), parent (sqpeer:1) UID (0)
EUID (0)
The kernel is 2.4.20, but I'm afraid I do not know the version of grsec as the distro maintainer is on holiday.

I am a complete newbie with grsecurity, and I don't really understand the message beyond the suggestion of a core dump. Is grsec merely reporting a problem here, or causing one?
the above message is just reporting, but grsec restrictions may also be leading up to the conditions to cause it as well. it would help if you could post your kernel .config file (at least the grsec options) and also check your logs for messages from PaX (the above message often appears when PaX is killing a task but the core file limit is set to 0).
PaX Team
 
Posts: 2310
Joined: Mon Mar 18, 2002 4:35 pm

Postby nikj » Tue Jul 01, 2003 6:06 am

Thanks for this - informing me that it was merely a report led to me solving the problem, which was not grsec related at all.

The problem was that the raid driver was not being loaded on boot as it was missing from the initrd image. Once I edited the image to include it, all worked fine.

Cheers
nikj
 
Posts: 2
Joined: Fri Jun 27, 2003 3:36 am


Return to grsecurity support

cron