Process on VM is slow

[Cruicer]

I ran out of things to look at here, looking for some new ideas...here is run down of what I am facing...

Have a VM (VM1) 4gb of RAM, 2 vCPU.  Sitting on ESX host 1, NO SAN in play right now...totally local storage.

ESX Host 1 has 2 x 6 for a total of 12 logical Processors, ProLiant DL585 G6.

I had to rebuild ESX Host 1 to create a datastore w/ 2mb block size in order to fit a VM which has a disk larger then 256.  So I moved VM1 off onto a different ESX host.  Again all local storage so it was a shutdown / migrate.

After it moved, the processes this VM needs to do went extremely slow...and I'll get into the "process" here in a second.

So I figured it was because of where the VM was moved to...a more populated host with some CPU / Memory constraints.

So I rebuilt the ESX host and then moved this said VM back to it.  And the "process" is still running like crap.

The "process"

Production server at sight SG takes backups of SQL and WebSphere data, the robocopies it to a datadomain source.  The SG site DataDomain source then replicates that data to our DR site's datadomain device.  Once that is completed then the DR datadomain takes the data (and here is where the slowness comes into play) and unzips the data to the DR VM.  From that point out everything is slow on the DR VM.  Now prior to any of the relocation of the VM, the process of the DR datadomain device to the VM, some scripts ran locally on the VM to restore the SQL / WebSphere data blah blah blah would take maybe 2 hours soup to nuts.  NOW, I'd be lucky if it will complete in 15 hours at the rate it is running.

There are currently two VMs running on this host, and the other VM (VM2), the same Process ran with no issue, and it's the same process just different data, same DD device used, from the same geographical location.  The ESX host is not even close to running short on resouces but I really don't have any idea where to even start.

I don't think it's network since VM2 ran fine...using the same port group, same vmnic, same vSwitch on the same ESX Host.

I don't think it's the DD device since there are many other VMs that the same process described are using, and all others are fine.

The ONLY thing that is out of date are the VMtools...which I plan to upgrade as soon as I can.  But I really don't think my issue is there.

Anyone have anything...at this point ANY IDEA would be helpful, cause I am really at a lost here.

[a_p_]

If I understand this correctly the data is unzipped on local storage on the newly build ESX host? Does the RAID controller have BBWC/FBWC attached in order to operate in write-back mode or does it run in write-trough mode (without write-cache)? BBWC/FBWC makes a huge difference in disk performance.

André

[Cruicer]

Andre, I am guessing there is no easy way of telling other then booting the ESX host up into SmartStart and checking out the ACU correct as the server is at a different sight and a simply "crack the case" and look is not really an option.

[a_p_]

If you are lucky you will see the information whether there's a BBWC/FBWC attached to the RAID controller in the Health status.

You could also do a file copy test. Without write-cache you might see ~5-10MB/s compared to ~80-100 MB/s with BBWC/FBWC.

André

[Cruicer]

Doesn't appear to have it, but what I don't understand then is this...last week, this problem VM was on this exact same host, and ran with no issue...none what so ever.  But as soon as I moved it off, it got slow, and then as soon as I moved it back is stayed slow.  And there is another VM on this same host and it isn't have ANY issues.  That one is running fine.  Really at a lost here.  Nothing is making sense...