I was looking into some storage performance issues for our VMware environment and I'm seeing some odd output from ESXTOP in VMware. We are running ESXi 5.0 on Cisco UCS B200 M2 blades. We connect to our NetApp 3240 using NFS.
Below is an example of what we are seeing in ESXTOP. You'll see vmnic4 is receiving a ton of traffic packets, while the kernel port (vmk3) that is bound to vmnic4 is receiving a lot less. There is nothing else using vmnic4 on the host. vmnic5 is also on the same NFS vlan as vmnic4 and it is not receiving the traffic, so I don't think it's a broadcast storm or anything.
Has anyone else seen anything like this before?
We have about 15 hosts that are connected to this storage and I'm seeing the same behavior on all of them.
I would assume the PKTRX/s received should be about the same on the port (vmnic) as the vmkernel port (vmk3), assuming that port is used for nothing but the kernel port. Also, it is using it's own VLAN for our NFS storage traffic.
I would be glad to take a look, but I would need to see the complete ESXTOP output. I have a parsing program that breaks it out. I generally ask for the output to be captured for at least one day on 15 minute intervals. The output file can get rather large. I have an ftp site that it can be sent to, I will parse it, run it through some analysis and let you know i I find anything. If you are interested, I do this for free as part of my performance research.
Having at least one day of data allows me to see how the overall activity changes during the day and whether there is anything that is leading up to this event or any other anomalies that are occuring during the event.