ext3fs SAS 5/i raid

Tue Nov 28 10:03:14 CST 2006

How many physical drives do you have in the system? 

Also could you send the output of fdisk -l?

Thanks,
Patrick Boyd

-----Original Message-----
From: linux-poweredge-bounces at dell.com
[mailto:linux-poweredge-bounces at dell.com] On Behalf Of Steven Stringham
Sent: Tuesday, November 28, 2006 9:51 AM
To: Gordon Henderson; linux-poweredge-Lists
Subject: Re: ext3fs SAS 5/i raid

Gordon,
Thanks for your reply. That is what I thought.

As you can from the following log, I do have multiple drives in the
system. And I am only using one of them presently. When I look at the
drive partitions using Webmin - it only shows one.  This leads me to
believe that indeed - it is supposed to be hardware mirroring. Also,
when I go into the controller BIOS afterwards, it shows the drive
remirroring (after putting the drive back in).

[root at localhost misc]# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
                      450G  3.9G  423G   1% /
/dev/sda1              99M   17M   78M  18% /boot
tmpfs                1004M     0 1004M   0% /dev/shm

Nov 28 02:28:41 localhost kernel: SCSI subsystem initialized Nov 28
02:28:41 localhost kernel: Fusion MPT base driver 3.04.01 Nov 28
02:28:41 localhost kernel: Copyright (c) 1999-2005 LSI Logic Corporation
Nov 28 02:28:41 localhost kernel: Fusion MPT SAS Host driver 3.04.01 Nov
28 02:28:41 localhost kernel: ACPI: PCI Interrupt 0000:02:08.0[A] -> GSI
16 (level, low) -> IRQ 169 Nov 28 02:28:41 localhost kernel: mptbase:
Initiating ioc0 bringup Nov 28 02:28:41 localhost kernel: input: ImPS/2
Generic Wheel Mouse as
/class/input/input1
Nov 28 02:28:41 localhost kernel: ioc0: SAS1068:
Capabilities={Initiator} Nov 28 02:28:42 localhost kernel: scsi0 : ioc0:
LSISAS1068, FwRev=00063200h, Ports=1, MaxQ=511, IRQ=169
Nov 28 02:28:42 localhost kernel:   Vendor: ATA       Model: 
HDS725050KLA360   Rev: AB5A
Nov 28 02:28:42 localhost kernel:   Type:   
Direct-Access                      ANSI SCSI revision: 05
Nov 28 02:28:42 localhost kernel:   Vendor: ATA       Model: 
HDS725050KLA360   Rev: AB5A
Nov 28 02:28:42 localhost kernel:   Type:   
Direct-Access                      ANSI SCSI revision: 05
Nov 28 02:28:42 localhost kernel:   Vendor: Dell      Model: VIRTUAL 
DISK      Rev: 1028
Nov 28 02:28:42 localhost kernel:   Type:   
Direct-Access                      ANSI SCSI revision: 05
Nov 28 02:28:42 localhost kernel: SCSI device sda: 976562176 512-byte
hdwr sectors (500000 MB) Nov 28 02:28:43 localhost kernel: sda: Write
Protect is off Nov 28 02:28:43 localhost kernel: SCSI device sda: drive
cache: write through Nov 28 02:28:44 localhost kernel: SCSI device sda:
976562176 512-byte hdwr sectors (500000 MB) Nov 28 02:28:44 localhost
kernel: sda: Write Protect is off Nov 28 02:28:44 localhost kernel: SCSI
device sda: drive cache: write through Nov 28 02:28:44 localhost kernel:
sda: sda1 sda2 Nov 28 02:28:44 localhost kernel: sd 0:1:0:0: Attached
scsi disk sda Nov 28 02:28:45 localhost kernel: device-mapper: ioctl:
4.7.0-ioctl
(2006-06-24) initialised: dm-devel at redhat.com Nov 28 02:28:45 localhost
kernel: EXT3-fs: INFO: recovery required on readonly filesystem.
Nov 28 02:28:45 localhost kernel: EXT3-fs: write access will be enabled
during recovery.
Nov 28 02:28:45 localhost kernel: kjournald starting.  Commit interval 5
seconds Nov 28 02:28:45 localhost kernel: EXT3-fs: recovery complete.
Nov 28 02:28:45 localhost kernel: EXT3-fs: mounted filesystem with
ordered data mode.

Gordon Henderson wrote:
> On Tue, 28 Nov 2006, Steven Stringham wrote:
>
>> I recently got an 840 tower with the SAS 5/i raid controller, and hot

>> pluggable SATA drives. I configured the two drives to be mirrored in 
>> the controller. I then installed Fedora 6 x64 patched.
>>
>> I then proceeded to test the raid controller by yanking out a drive. 
>> I was expecting the system to maybe say something occurred. But, the 
>> ext3 file system complained about a journaling I/O error, and then 
>> remounted the file system in Read Only mode.
>>
>> I called tech support, and they said that this was normal for all 
>> RAID controllers and ext3.
>>
>> So, after all this my question is: If I am using hardware RAID, 
>> should the linux kernel remount the file system to readonly when a 
>> mirrored drive fails? This seems very counterintuitive to me. Should 
>> the system not continue to function, but with some notification of 
>> the hardware failure?
>
> Not replying to the list as my experiences aren't directly related to 
> Dells hardware solutions, but I have to say if thats the behaviour of 
> the system then I'd throw it back at them.
>
> I have built dozens of systems using Linux s/w RAID with mirrored 
> drives and I can yank a drive and (after a brief pause when it 
> sometimes does a bus reset) the system jsut carried on - which is what

> the whole point of RAID is about!
>
> ext3 can be configured to remount in read-only mode, cause a panic or 
> ignore any underlying filesystem errors - thats an option in the 
> /etc/fstab (read only is the default in later kernels IIRC) but 
> really, yankking a drive in a correctly configured system should not 
> have any impact on the OS at all.
>
> Are you sure you uses mirroring and not striping? What is the total 
> capacity of your system vs. the capacity of each disk? With mirroring,

> the capacity ought to be the same (or marginally smaller) than the 
> capacity of each disk. With striping you'll see double the capacity.
> (and half the redundancy)
>
> Gordon

_______________________________________________
Linux-PowerEdge mailing list
Linux-PowerEdge at dell.com
http://lists.us.dell.com/mailman/listinfo/linux-poweredge
Please read the FAQ at http://lists.us.dell.com/faq