I want to make sure I understand what I am seeing when I look at esxtop in interactive mode. I guess a better way to say it is I want to make sure that my interpetation is correct..
PORT-ID USED-BY TEAM-PNIC DNAME PKTTX/s MbTX/s PKTRX/s MbRX/s %DRPTX %DRPRX
33554434 vmnic2 - vSwitch1 640.87 3.35 725.51 1.14 0.00 0.00
33554435 vmnic3 - vSwitch1 608.21 4.54 833.51 3.25 0.00 0.00
33554436 vmnic4 - vSwitch1 914.10 6.04 318.29 0.34 0.00 0.00
33554439 14414:test1i_L45 all(3) vSwitch1 1166.58 10.60 840.90 1.39 0.00 0.03
33554440 14414:test1_L45 all(3) vSwitch1 0.00 0.00 0.00 0.00 0.00 100.00
33554479 28611012:test2_L all(3) vSwitch1 177.15 1.57 118.49 0.39 0.00 0.20
33554480 28611012:test2_L all(3) vSwitch1 161.65 0.39 220.54 1.67 0.00 0.11
I am seeing %DRPRX value get to 100% for both vm's. However there are 2 line entries for each vm in this output. I assume this is because there are 2 port ids as each vm belongs to 2 seperate tagged vlans. This requires each vm to belong to 2 port groups... (not sure on this part)
What I am trying to understand is why there are higher percentage drops showing on one of the 2 entries for a given host that shows little throughput. There is still small percentage of drops across the other interfaces but technically this is all in memory so is it not the same buffers??? I have found articles that talk about increaseing rx buffers or chaning the vnic driver to vmxnet3 (this is solaris guest os).
These drop percentages are constant so before I make any changes I want to make sure I am reading this correctly and understang it. There is much about the internals of vmware that I am still learning.
Thanks in advance for any assistance in understanding this data better.