• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 743
  • Last Modified:

Windows 2008 with RDM in ESXi 4.1 running exchange 2007 slow to copy 1 gig file to same drive

Hi ,

We have a windows 2008 sp2 server with Exchange 2007 installed with the datastore on a RDM attached directly within Vmware.

Server runs fine but we have had issues with slow backups. After looking into it we found that copying large files onto the same drive i.e. C:\ the file copy would take a very long time the first time (over 5 minutes for a 1 gig file ) after that it would be okay.

Any ideas?

0
pancho15
Asked:
pancho15
  • 8
  • 8
1 Solution
 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
what underlying datastore is the C drive virtual disk on?

e.g. RAID type, SAS or SATA, number of disks, speed of disks etc?
0
 
pancho15Author Commented:
Disk resides on an equallogic SAN ( 10 gig san with 600 gig sas drives in raid 5 , which is not the problem. Many other servers reside on the same datastore without issues.
0
 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
virtual servers? RAID 5 is not particulary fast!
0
Industry Leaders: We Want Your Opinion!

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Check write and read speed benchmarks, with Crystaldisk Benchmark and upload here.
0
 
pancho15Author Commented:
Raid 5 on SAN is recommeneded. They are pretty Dam quick . they also use a 10gig network for access.

I dont believe it's an issue with the disks as all other servers sit on the same environment without a problem. It's seems like its an issue or bug in 2008 - being looking for a hotfix but had not luck for my specific problem.

I will run the benchmark tool but again I'm 100% it's not the issue
0
 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
the bechmark tool will give me some benchmarks to compare!
0
 
pancho15Author Commented:
Hanccocka,

I know I said I was 100% it wasnt the SAN, but I think I'm wrong!!!

Vmware run the esxtop - D tool and advised that when running a backup or a disk copy action the DAVG level should not exceed 20. But in our case it was exceeding 200!!.

We are using the Dell Equallogic PS6010 SAN which is a 15 Drive 10G network ISCSI SAN ESXTOP - D
0
 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
0
 
pancho15Author Commented:
Crystal Mark results below !! not looking to good

Physical Server with 15k RPM drives in Raid 5 (Dell 2950) 500MB test
SEQ= 314 Read   248 Write
512K= 74.49 Read   80 Write
4k = 1.4 Read 2.4 Write
4k QD= 8.4  read  2.6 Write

VM 2003 Server 500MB test
SEQ= 8.4 Read   201.7 Write
512K= 3.8 Read   155 Write
4k = 1.0 Read 8.4 Write
4k QD= 25.7  read  32.1 Write

VM 2008 Server 500MB test
SEQ= 5.0 Read   119.8 Write
512K= 3.4 Read   70.9 Write
4k = 0.9 Read 6.2 Write
4k QD= 10.8  read  10.86 Write



0
 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
Can you change the Read and Write Cache for the SAN?
0
 
pancho15Author Commented:
Will check this morning and let you know. I also have a case open with dell equallogic , so I will provide these results.
0
 
pancho15Author Commented:
Update:

After changing the MTU value to 1500 on each of the vswitches used by the ISCSI networks I'm now seeing much better performance. My Symantec backupexec remote agent is now backing up the VM at 3400 MB/min . This is treating the VM as a physical machine not using the Vmware agent.

CrystalDisk Mark
Seq Read 111.4 MB and 102 MB write

Previous results were 8.4 MB Read and 201 Write

So performance is still down and I'm unsure what would cause such a slow down when using Jumbo Frames.

I've check everything to ensure JF is enabled end to end. dont know what else to check
0
 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
What were the MTUs on the vSwitches?

1500 is the default?

were you attempting to use Jumbo Frames, if you were Jumbo Frames needs to be enabled on all the network, and SAN. (and ESXi).
0
 
pancho15Author Commented:
ESXi Server 1

VSwitch3 MTU 9000 Uplinks vmnic3, vmnic2

vmnic3 MTU 9000
vmnic2 MTU 9000

Equallogic is set for Jumbo frames and so is the PowerConnect 8024F, not sure what else to check

IS my storage performance still slow read of 114MB and 102 MB write?
0
 
Andrew Hancock (VMware vExpert / EE MVE^2)VMware and Virtualization ConsultantCommented:
So, there was a mismatched, jumbo frame somewhere?

But I must say, although jumbo frames are recommended, it works for some and not others!

ah, yes it's still slow.

just as an example of 5,400 rpm SATA disks, not on a particular fast system

two disks - 84.5 MB/s Write,  150 MB/s Read
four disks - 196 MB/s Write, 276 MB/s Read
eight disks - 212 MB/s Write,  287 MB/s Read

So compared to my "noddy - cheap system" it's slow!
0
 
pancho15Author Commented:
thanks - still working with Dell to resolve slowness issues
0

Featured Post

Free Tool: Site Down Detector

Helpful to verify reports of your own downtime, or to double check a downed website you are trying to access.

One of a set of tools we are providing to everyone as a way of saying thank you for being a part of the community.

  • 8
  • 8
Tackle projects and never again get stuck behind a technical roadblock.
Join Now