[CentOS] Poor RAID performance new Xeon server?

Stewart Williams lists at pinkyboots.co.uk
Sun Jan 11 21:12:01 UTC 2009


Stewart Williams wrote:
> John R Pierce wrote:
>> Stewart Williams wrote:
>>> I have just purchased an HP ProLiant HP ML110 G5 server and install ed 
>>> CentOS 5.2 x86_64 on it.
>>>
>>> It has the following spec:
>>>
>>> Intel(R) Xeon(R) CPU 3065  @ 2.33GHz
>>> 4GB ECC memory
>>> 4 x 250GB SATA hard disks running at 1.5GB/s
>>>
>>> Onboard RAID controller is enabled but at the moment I have used mdadm 
>>> to configure the array.
>>>
>>> RAID bus controller: Intel Corporation 82801 SATA RAID Controller
>>>
>> that is essentially desktop grade disk IO
>>
>>
>>> For a simple striped array I ran:
>>>
>>> # mdadm --create /dev/md0 --level=0 --raid-devices=2 /dev/sdb1 /dev/sdc1
>>> # mke2fs -j /dev/md0
>>> # mount -t ext3 /dev/md0 /mnt
>>>
>>> Attached are the results of 2 bonnie++ tests I made to test the 
>>> performance:
>>>
>>> # bonnie++ -s 256m -d /mnt -u 0 -r 0
>>>
>>> and
>>>
>>> # bonnie++ -s 1g -d /mnt -u 0 -r 0
>>>
>>> I also tried 3 of the drives in a RAID 5 setup with gave similar results.
>>>
>>> Is it me or are the results poor?
>>>
>>> Is this the best I can expect from the hardware or is something wrong?
>>>
>>> I would appreciate any advice or possible tweaks I can make to the 
>>> system to make the performance better.
>>>
>>> The block I/O is the thing that concerns me as mostly I am serving a 
>>> 650MB file via samba to 5 clients and I think this is where I need the 
>>> speed.
>> is this a sequential or random access application thats using this 
>> file?   is it read only/mostly, or is it random update?
> 
> I'm not sure, how can I find this out?
> 
>> its rather hard to read your bonnie output logs as they aren't very 
>> columnar. but it appears the sequetial read speed at least is really high.
>>
>> i'm seeing 55MB/sec random(block) and 1.4GB/sec sequential reads on the 
>> 1GB file,
> 
> Correct.
> 
>> so I dunno what your issues are...   of course, a 1GB file 
>> sits entirely in the system cache assuming a reasonable amount of 
>> otherwise idle memory
> 
> I'm not sure whether the performance would suffice as I've not tried 
> putting it in production.
> 
> I am going to benchmark the old server (currently in production) that 
> this is replacing.
> 
> Thanks,
> 
> Stewart

I've ran the same bonnie++ test on my old server using a 1GB file.

The machine has only 1GB ram and 2 IDE ATA100 hard disks in a RAID 1 
mirror on seperate IDE channels on the motherboard.

I got about 38MB/s write w/ 13% CPU and  80MB/s read w/ 97% CPU. Also if 
I watch `top` with only one user, and run a quick-report in quickbooks 
on a stock item, the iowait is about 50% and the cached ram fills to 
around 200-300MB. So with 5 users I expect it that to go up quite 
dramatically.


More information about the CentOS mailing list