Hoiw about some results from a VM? Those are freakin smokin numbers.

I was going to test MS iSCSI initiator from within VM to compare that to the test to vmdk using the ESX SW iSCSI. To do that can I just add a Virtual Machine Port Group to the existing switch which runs the VMkernel and SC (connecting back to iSCSI network) or should I create a separate virtual switch with just a VM Port Group back to that same iSCSI network?

Tested this on an very utilized SAN (HP EVA5000 with VCS 3.028 active/passive).
The SAN contains 125 LUNs used by 50 physical servers.
14 of those physical servers are ESX servers hosting 90 virtual servers using the same SAN.
It is a mixed usage. Dev, Test, QA and Prod.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
TABLE OF RESULTS
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

SERVER TYPE: Win2003 VM ON ESX 3.0.2 (On a host with 7 other active VMs)
CPU TYPE / NUMBER: VCPU / 1
HOST TYPE: HP ProLiant BL25p G1 , 16 GB RAM, 2x 2.4 GHz AMD Opteron 280 (Dual core)
STORAGE TYPE / DISK NUMBER / RAID LEVEL: HP EVA 5000 / 68 X 300 GB FC 10k / Vraid5
SAN TYPE / HBAs : 2 GB FC, QLogic QLA2312 HBA
VMFS: 150 GB LUN, 1MB Block Size

##################################################################################
TEST NAME-------------------Av. Resp. Time ms------Av. IOs/sek-------Av. MB/sek------
##################################################################################

Max Throughput-100%Read........____11____..........___4892___.........___153____

RealLife-60%Rand-65%Read......____13____..........___3631___.........____28____

Max Throughput-50%Read..........____29____..........___1911___.........____60___

Random-8k-70%Read.................____13____..........___3640___.........____28____

EXCEPTIONS: VCPU Util. 57-59-42-60 %;

##################################################################################

Yeah its pretty good, but remember its not raid 5 its 1+0...

I can get some numbers soon on VM's on both raid 5 sets and 1+0, the same san also has some SATA storage i can possible test that as well if i get the time

There will definately be a lot of cache hits, but the idea of testing is to test the whole san, cache included, i should also try enabling the write cache on the physical disks....

The VM's that i have running can boot windows 2003 in about 9 seconds from VM POST to the alt+ctrl+del logon screen

Remark: 16 Active VM's on same SAN LUN.

[code]
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
TABLE OF RESULTS
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

SERVER TYPE: Win2003 VM ON ESX 3.0.1
CPU TYPE / NUMBER: VCPU / 1
HOST TYPE: HP ProLiant DL585G2 , 20 GB RAM, 4 x Dual-Core AMD Opteron(tm) Processor 8216 2411
STORAGE TYPE / DISK NUMBER / RAID LEVEL: Sun Storagetek Flexline 380 / 9 X 70 GB FC 10k / raid5
SAN TYPE / HBAs : 2 GB FC, QLogic QLA2312 HBA / 2 x Brocade 24K
VMFS: 500 GB LUN, 1MB Block Size

##################################################################################
TEST NAME-------------------Av. Resp. Time ms------Av. IOs/sek-------Av. MB/sek---
##################################################################################

Max Throughput-100%Read........____9____..........___5785___.........____180___

RealLife-60%Rand-65%Read.......____36___..........___1421___.........____11____

Max Throughput-50%Read.........____7____..........___6404___.........____210___

Random-8k-70%Read..............____33___..........___1549___.........____12____

##################################################################################
[/code]

Remark: Same VM migrated to older server and tested on local storage with no raid.

[code]
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
TABLE OF RESULTS
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

SERVER TYPE: Win2003 VM ON ESX 3.0.1
CPU TYPE / NUMBER: VCPU / 1
HOST TYPE: HP ProLiant DL580G2 , 20 GB RAM, 4 x Intel(R) Xeon(TM) MP CPU 2.80GHz 2794
STORAGE TYPE / DISK NUMBER / RAID LEVEL: 1 X 300 GB U320 SCSI RAID0
SAN TYPE / HBAs : Local SmartArray 5i
VMFS: 300GB LUN, 1MB Block Size

##################################################################################
TEST NAME-------------------Av. Resp. Time ms------Av. IOs/sek-------Av. MB/sek---
##################################################################################

Max Throughput-100%Read........____36___..........___1616___.........____50____

RealLife-60%Rand-65%Read.......____53___..........____973___........._____7____

Max Throughput-50%Read.........____21___..........___2639___.........____82____

Random-8k-70%Read..............____58___..........____887___........._____7____

##################################################################################
[/code]

I think I may be the first to post with some EMC numbers

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
TABLE oF RESULTS
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

SERVER TYPE: Win2k3 VM
CPU TYPE / NUMBER: VCPU / 1
HOST TYPE: Dell PE2900, 24GB RAM; 2x XEON 5160, 3.00 GHz, DC
STORAGE TYPE / DISK NUMBER / RAID LEVEL: EMC CX500 x 1 / 7 FC(10k)Disks / R5

##################################################################################
TEST NAME-------------------Av. Resp. Time ms------Av. IOs/sek-------Av. MB/sek------
##################################################################################

Max Throughput-100%Read........___9.98___..........___5846___.........___182____

RealLife-60%Rand-65%Read......___33.29__..........___1486___.........___11.61__

Max Throughput-50%Read..........___8.33___..........___6889___.........___215____

Random-8k-70%Read.................___29.13__..........___1628___.........___12.72___

EXCEPTIONS: CPU Util.-54,38,62,41;

##################################################################################