Why do I get Dirty FS after fsck on an Adaptec RAID 3405 RAID1?

Question

SUSE

Good afternoon community. The machine is running OpenSUSE:

oraprod:~ # uname -a 
Linux oraprod 2.6.31.12-0.2-desktop #1 SMP PREEMPT 2010-03-16 21:25:39 +0100 x86_64 x86_64 x86_64 GNU/Linux 
oraprod:~ # cat /etc/*release* 
openSUSE 11.2 (x86_64) 
VERSION = 11.2

The / partition is on a RAID1 mirror on an Adaptec RAID 3405 controller. Once after a reboot, the partition was mounted in RO with the Dirty flag. We decided to run fsck, which succeeded. But after the reboot, the section again “Filesystem still has errors”.
Smart hard drives turned out to be normal (At least, the data that the controller gives about the drives). Hence the question: who can fool - the controller or the disks? Why don't other arrays on the same controllers fool around like that? (there are a few more RAID10s out there). If disks have flown why both are synchronous and spot-check does not fall apart? Ready to provide missing data upon request.

Answer 1

An interesting detail emerged. On Linux, with an RW-mounted file system, fsck -n will show errors. And I, brought up on FreeBSD, simply did not know this, because. there were no such fsck kicks.