Undocumented requirement for dvswitch VLAN health check
I really welcome the dvswitch health checks as we in the past have experienced problems with VLAN's not available on all the configured physical switchports, and theese misconfigurations are not always easy to find. We have enabled the health checks on a few clusters, but there are at least one requirement for the VLAN-monitoring that not all our clusters meet.
The dvswitch health checks can monitor the following stuff once per minute (default):
- VLAN – Checks whether vSphere distributed switch VLAN settings match trunk port configuration on the adjacent physical switch ports
- MTU – Checks whether the physical access switch port MTU setting based on per VLAN matches the vSphere distributed switch MTU setting
- Network adapter teaming – Checks whether the physical access switch ports EtherChannel setting matches the distributed switch distributed port group IP Hash teaming policy settings.
Some limitations of the health checks are documented in KB#2034795, but I think I found one more.
So, what's wrong here? On the cluster where we only have 1GbE-ports we normally use two uplinks for management and two uplinks for VM traffic. In this case we have the management- an VMotion-VLANs on uplink2 and uplink4 while the VM traffic is on uplink1 and uplink3.
As far as I can see the VLAN health check do assume that all VLAN's are present on all uplinks. I've sent my comments to KB#2034795, so let's see if VMware can give me an answer.
On our newer clusters where we use 10GbE and have all VLAN's present on all uplinks this is of course not an isseue and the health check works as we expected!