[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-users] XCP 1.6 mdadm


  • To: <xen-users@xxxxxxxxxxxxx>
  • From: <Frank@xxxxxxxxxxxxxxx>
  • Date: Sun, 2 Jun 2013 12:27:50 +0100
  • Delivery-date: Sun, 02 Jun 2013 11:29:16 +0000
  • List-id: Xen user discussion <xen-users.lists.xen.org>
  • Thread-index: Ac5fZwqMQJdovbgmSxiUvNdPjdq2uA==
  • Thread-topic: XCP 1.6 mdadm

Hello,
Can anyone tell me why virtual disk performance on an lvm local sr
running on mdadm raid 1 should be very slow?  Please see test results
below.


System 1
XCP 1.6 running on mdadm raid1

fdisk -l output

----------------------------------------------
Disk /dev/sda: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1          13      104391   83  Linux
/dev/sda2              14         654     5148832+  fd  Linux raid
autodetect
/dev/sda3             655       13690   104711670   fd  Linux raid
autodetect

Disk /dev/sdb: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1   *           1          13      104391   83  Linux
/dev/sdb2              14         654     5148832+  fd  Linux raid
autodetect
/dev/sdb3             655       13690   104711670   fd  Linux raid
autodetect

Disk /dev/md0: 5272 MB, 5272305664 bytes
2 heads, 4 sectors/track, 1287184 cylinders
Units = cylinders of 8 * 512 = 4096 bytes

Disk /dev/md0 doesn't contain a valid partition table

Disk /dev/md1: 107.2 GB, 107224629248 bytes
2 heads, 4 sectors/track, 26177888 cylinders
Units = cylinders of 8 * 512 = 4096 bytes

Disk /dev/md1 doesn't contain a valid partition table
-------------------------------------------------

sda1 and sdb1 are both 100M boot partitions

XCP running on /dev/md0 (sda2 + sdb2 raid1)
LVM SR on /dev/md1 (sda3 + sdb3 raid1)

Benchmarks are run using
dd bs=1M count=512 if=/dev/zero of=test conv=fdatasync

Benchmark 1 run in / (xcp system root) = 53.1 MB/s
Benchmark 2 run in /sbb/test (10G logical volume on Local Storage
formatted ext3 and mounted on /sbb/test =  94.1 MB/s
Benchmark 3 run in root of centos 5 fully paravirtualised vm on local
storage = 31.5 MB/s

--------------------------------------------------

System 2
Single disk (USING ONE OF THE ABOVE DISKS)
FRESH STANDARD INSTALLATION OF XCP 1.6 WITH LVM LOCAL STORAGE

Standard GPT partitioning

When the above benchmarks are run on this system the results are as
follows:

Benchmark 1 run in / (xcp system root) = 89.5 MB/s
Benchmark 2 run in /sbb/test (10G logical volume on Local Storage
formatted ext3 and mounted on /sbb/test =  92.0 MB/s    
Benchmark 3 run in root of centos 5 fully paravirtualised vm on local
storage = 61.0 MB/s  (Same vm as above imported from xva file)

------------------------------------------------------------------------
------------------------------------

My questions are:
1. Why is benchmark 1 slower on mdadm raid 1?
2  Why is benchmark 3 slower on mdadm raid 1?
Yet benchmark 2 results are roughly the same.
Am I missing something here and is it possible to tweak the performance
of XCP on raid?

Regards,
Frank.

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.