Hi,
I have 2 vSphere hosts with latest patches connected to a HP EVA 4400 via fibre. I have had to move some VM's from one LUN to another. Both LUN's are on FSAS disks.
The problem is they are taking an age to move...
I started looking at diagnostics during a storage vMotion and ran the following:
esxtop
d
f
b,c,d,e,h,j
s,2
For a while it stays around these levels:
2:32:05pm up 28 days 2:21, 175 worlds; CPU load average: 0.14, 0.15, 0.14
ADAPTR NWDS CMDS/s READS/s WRITES/s MBREAD/s MBWRTN/s DAVG/cmd KAVG/cmd GA
vmhba1 182 2166.27 2079.01 87.26 16.24 35.52 0.33 0.00
vmhba2 182 4.77 1.43 3.34 0.02 0.02 0.22 0.01
vmhba3 3 27.66 0.00 27.66 0.00 0.17 0.05 0.00
Then after a bit the CMDS/s drops to anywhere from 10 - 50, however in the GUI all it shows is the following:
18% Migrating the active state of the virtual machine.
As an example of speed it has taken 1 hour 30 minutes to move a thin provisioned VM, it has a provisioned size of 31GB (30GB Hdd + 1 GB RAM) with a used size of 5.69GB..., so thats an hour and a half to move just under 6GB + RAM?
The strange thing is another VM which was thick provisioned at 41GB (40GB disk + 1GB RAM) took only 1 hour.
I have monitored the storage system during these moves and it does not seem to be heavily utilized (thrashing etc) so I am wondering where the delay is coming from. I would appreciate any help.