DL 360 G4 system crash under SLES9 "ASR detected by system ROM"

[Juan B.S.R.]

Hello.
I am having problems with one HP DL360 G4 under SuSe Linux Enterprise Server 9.
Sometimes system crashes and the Integrated Management Log show "ASR Detected by system ROM".
After reboot I can see the following message in Logs "Unable to handle kernel paging request at virtual address 0121f083". In the bottom of this message you can see the error in more detail and in the attached file you can see more errors like this, but in another processes (kswapd, nfsd, vbda).
I dont know if this is a hardware or software error. Maybe there are a defective memory module or a kernel bug.
Thanks for all and best regards

-------------
Full error report

Feb 9 19:09:35 leda kernel: Unable to handle kernel paging request at virtual address 0121f083
Feb 9 19:09:35 leda kernel: printing eip:
Feb 9 19:09:35 leda kernel: c0150135
Feb 9 19:09:35 leda kernel: *pde = 00000000
Feb 9 19:09:35 leda kernel: Oops: 0002 [#1]
Feb 9 19:09:35 leda kernel: SMP
Feb 9 19:09:35 leda kernel: CPU: 0
Feb 9 19:09:35 leda kernel: EIP: 0060:[] Tainted: PF U
Feb 9 19:09:35 leda kernel: EFLAGS: 00010046 (2.6.5-7.97-smp)
Feb 9 19:09:35 leda kernel: EIP is at cache_alloc_refill+0x125/0x5d0
Feb 9 19:09:35 leda kernel: eax: 0121f07f ebx: e7ccf000 ecx: ffffffff edx: f7e4499c
Feb 9 19:09:35 leda kernel: esi: 00000006 edi: e7ccf018 ebp: f7997280 esp: caf87d18
Feb 9 19:09:35 leda kernel: ds: 007b es: 007b ss: 0068
Feb 9 19:09:35 leda kernel: Process vbda (pid: 30408, threadinfo=caf86000 task=f1d381c0)
Feb 9 19:09:35 leda kernel: Stack: f7997290 f7e4499c 0863c080 00000000 f7e449a4 000000d0 f7e44914 f7e44914
Feb 9 19:09:35 leda kernel: 00000296 caf87d94 f7928400 c01506a4 00000000 f7928400 f920b99f c0185f55
Feb 9 19:09:35 leda kernel: 00000000 f7928400 caf87d94 cde141b8 c01863b0 caf87d94 de58e6f0 f91fd320
Feb 9 19:09:35 leda kernel: Call Trace:
Feb 9 19:09:35 leda kernel: [] kmem_cache_alloc+0x44/0x50
Feb 9 19:09:35 leda kernel: [] reiserfs_alloc_inode+0xf/0x20 [reiserfs]
Feb 9 19:09:35 leda kernel: [] alloc_inode+0x25/0x1a0
Feb 9 19:09:35 leda kernel: [] iget5_locked+0xc0/0x1a0
Feb 9 19:09:35 leda kernel: [] reiserfs_find_actor+0x0/0x30 [reiserfs]
Feb 9 19:09:35 leda kernel: [] reiserfs_iget+0x37/0xa0 [reiserfs]
Feb 9 19:09:35 leda kernel: [] reiserfs_init_locked_inode+0x0/0x10 [reiserfs]
Feb 9 19:09:35 leda kernel: [] reiserfs_lookup+0x16e/0x200 [reiserfs]
Feb 9 19:09:35 leda kernel: [] d_alloc+0x1b/0x1c0
Feb 9 19:09:35 leda kernel: [] real_lookup+0xbe/0xf0
Feb 9 19:09:35 leda kernel: [] do_lookup+0x7e/0x90
Feb 9 19:09:35 leda kernel: [] link_path_walk+0x5b0/0xa90
Feb 9 19:09:35 leda kernel: [] getname+0x83/0xb0
Feb 9 19:09:35 leda kernel: [] __user_walk_it+0x59/0x70
Feb 9 19:09:35 leda kernel: [] vfs_lstat+0x65/0x140
Feb 9 19:09:35 leda kernel: [] sys_lstat64+0xf/0x30
Feb 9 19:09:35 leda kernel: [] sys_alarm+0x2a/0x50
Feb 9 19:09:35 leda kernel: [] sysenter_past_esp+0x52/0x79
Feb 9 19:09:35 leda kernel:
Feb 9 19:09:35 leda kernel: Code: 89 50 04 89 02 83 7b 14 ff c7 03 00 01 10 00 c7 43 04 00 02

[Anthony Martin_1]

Hi Juan,
What version of the Insight Management Agents do you have loaded on the server.
I suggest that you stop the HP Insight Agent processes. These monitor the server to see if it is "hung", if so, it performs an ASR and the server reboots.

Are you able to ping the server before it crashes.
More information is required please.

Cheers
Anthony

[Juan B.S.R.]

Hi Anthony.
The version of the "hp System Health Application and Insight Management Agents" is 7.3.0-58.
Before ASR the server really seems to be hung and rebooted correctly.

[Anthony Martin_1]

Hi Juan,
A couple of questions
Were you doing any backups/restores at the time (what is your backup software). Data Protector has a vbda process, and that has been known to have problems (patches are available).
Do you use Xen?
The EIP suggests that the function that crashed the system was "cache_alloc_refill".

Just trying to get more info.

Anthony

[Juan B.S.R.]

Hello.
The backup software is Data Protector 5.10. and it was making backup.
At that moment the crashed process was "vbda" but in other time was "nfsd" or "kswapd". In the first post there are an attached file with all the last errors.
I´m not using Xen. Why?

[Anthony Martin_1]

Juan,
Have you applied the patches to DataProtector. It looks like this is where your problem is coming from.

[Juan B.S.R.]

Hi.Anthony.
I think that "Data Protector" is not the problem, however I will go to patch it.
I think the problem is something about heavy I/O transfer to disk.
Also I have disabled ACPI in GRUB boot parameters.