Re: Ceph performance with 8K blocks.

"Campbell, Bill" <bcampbell@xxxxxxxxxxxxxxxxxxxx> · Tue, 17 Sep 2013 10:56:09 -0400 (EDT)

Windows default (NTFS) is a 4k block.  Are you changing the allocation unit to 8k as a default for your configuration?

From: "Gregory Farnum" <greg@xxxxxxxxxxx>
To: "Jason Villalta" <jason@xxxxxxxxxxxx>
Cc: ceph-users@xxxxxxxxxxxxxx
Sent: Tuesday, September 17, 2013 10:40:09 AM
Subject: Re:  Ceph performance with 8K blocks.

Your 8k-block dd test is not nearly the same as your 8k-block rados bench or SQL tests. Both rados bench and SQL require the write to be committed to disk before moving on to the next one; dd is simply writing into the page cache. So you're not going to get 460 or even 273MB/s with sync 8k writes regardless of your settings.

However, I think you should be able to tune your OSDs into somewhat better numbers -- that rados bench is giving you ~300IOPs on every OSD (with a small pipeline!), and an SSD-based daemon should be going faster. What kind of logging are you running with and what configs have you set?

Hopefully you can get Mark or Sam or somebody who's done some performance tuning to offer some tips as well. :)
-Greg

On Tuesday, September 17, 2013, Jason Villalta  wrote:

Hello all,  I am new to the list.

I have a single machines setup for testing Ceph.  It has a dual proc 6 cores(12core total) for CPU and 128GB of RAM.  I also have 3 Intel 520 240GB SSDs and an OSD setup on each disk with the OSD and Journal in separate partitions formatted with ext4.

My goal here is to prove just how fast Ceph can go and what kind of performance to expect when using it as a back-end storage for virtual machines mostly windows.  I would also like to try to understand how it will scale IO by removing one disk of the three and doing the benchmark tests.  But that is secondary.  So far here are my results.  I am aware this is all sequential, I just want to know how fast it can go.

DD IO test of SSD disks:  I am testing 8K blocks since that is the default block size of windows.
 dd of=ddbenchfile if=/dev/zero bs=8K count=1000000
8192000000 bytes (8.2 GB) copied, 17.7953 s, 460 MB/s

dd if=ddbenchfile of=/dev/null bs=8K
8192000000 bytes (8.2 GB) copied, 2.94287 s, 2.8 GB/s

RADOS bench test with 3 SSD disks and 4MB object size(Default):

rados --no-cleanup bench -p pbench 30 write
Total writes made:      2061
Write size:             4194304
Bandwidth (MB/sec):     273.004

Stddev Bandwidth:       67.5237

Max bandwidth (MB/sec): 352
Min bandwidth (MB/sec): 0
Average Latency:        0.234199
Stddev Latency:         0.130874
Max latency:            0.867119
Min latency:            0.039318

-----
rados bench -p pbench 30 seq
Total reads made:     2061
Read size:            4194304
Bandwidth (MB/sec):    956.466

Average Latency:       0.0666347

Max latency:           0.208986
Min latency:           0.011625

This all looks like I would expect from using three disks.  The problems appear to come with the 8K blocks/object size.

RADOS bench test with 3 SSD disks and 8K object size(8K blocks):
rados --no-cleanup bench -b 8192 -p pbench 30 write
Total writes made:      13770
Write size:             8192

Bandwidth (MB/sec):     3.581

Stddev Bandwidth:       1.04405
Max bandwidth (MB/sec): 6.19531
Min bandwidth (MB/sec): 0
Average Latency:        0.0348977

Stddev Latency:         0.0349212
Max latency:            0.326429
Min latency:            0.0019
------
rados bench -b 8192 -p pbench 30 seq
Total reads made:     13770

Read size:            8192
Bandwidth (MB/sec):    52.573

Average Latency:       0.00237483
Max latency:           0.006783
Min latency:           0.000521

So are these performance correct or is this something I missed with the testing procedure?  The RADOS bench number with 8K block size are the same we see when testing performance in an VM with SQLIO.  Does anyone know of any configure changes that are needed to get the Ceph performance closer to native performance with 8K blocks?

Thanks in advance.

-- 
-- 

Jason Villalta
Co-founder

800.799.4407x1230 | www.RubixTechnology.com

-- 
Software Engineer #42 @ http://inktank.com | http://ceph.com

_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

NOTICE: Protect the information in this message in accordance with the company's security policies. If you received this message in error, immediately notify the sender and destroy all copies.

_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com