Megaraid problems EL4 U5 64 (Ora with direct IO + async IO)

[Rainer Traut]

Hi,

we have:
PE1850, PERC 4e/Si
Redhat EL4 U5, Kernel 2.6.9-55.0.9.ELsmp
Ora10G R10.1.0.5.0 64bit

Megaraid drivers:
megaraid cmm: 2.20.2.6rh (Release Date: Tue Jan 16 12:35:06 PST 2007)
megaraid: 2.20.4.6-rh2 (Release Date: Wed Jun 28 12:27:22 EST 2006)

PERC FW: 522D

Oracle has async IO as well as direct IO enabled.

We run regularly RMAN backups and see the server then losing access to 
the Raid device.
There is nothing in the ESM Log but syslog via network gives this:

Oct  9 18:05:10 neptun syslogd: /var/log/messages: Read-only file system
Oct  9 18:05:10 neptun kernel: megaraid: aborting-3345892 cmd=28 <c=1 
t=0 l=0>
Oct  9 18:05:10 neptun kernel: megaraid abort: 3345892:14[255:128], fw 
owner
... and repeats...
Oct  9 18:05:14 neptun kernel: EXT3-fs error (device dm-0) in 
start_transaction: Journal has aborted
Oct  9 18:05:14 neptun syslogd: /var/log/maillog: Read-only file system
Oct  9 18:05:14 neptun kernel: EXT3-fs error (device dm-0) in 
start_transaction: Journal has aborted
Oct  9 18:05:20 neptun kernel: scsi2 (0:0): rejecting I/O to offline device
Oct  9 18:05:20 neptun kernel: EXT3-fs error (device dm-0): 
ext3_find_entry: reading directory #2097799 offset 0
Oct  9 18:05:20 neptun kernel:
Oct  9 18:05:20 neptun kernel: scsi2 (0:0): rejecting I/O to offline device
Oct  9 18:05:20 neptun kernel: EXT3-fs error (device dm-0): 
ext3_find_entry: reading directory #2097876 offset 0
Oct  9 18:05:20 neptun kernel:
Oct  9 18:05:20 neptun kernel: scsi2 (0:0): rejecting I/O to offline device
Oct  9 18:05:20 neptun kernel: EXT3-fs error (device dm-0): 
ext3_find_entry: reading directory #2061299 offset 0
Oct  9 18:05:20 neptun kernel:

...

Anybody seen this?

Rainer