What would cause a vCMP Management interface to go unitialized, but the subinterfaces under that management interface were still up? I did not find any logs that were helpful and ended up rebooting the vCMP in question. Any ideas on what might have caused this? Has anyone seen this before? It appeared the management did not see the cable connected to the interfaces....See screenshot:
Excellent. That is what I thought. What you are seeing may be by design. Please see the following:
Namely, the following is important: "While each blade is assigned an individual management IP address used to access the blade directly, a floating management IP address is also present to facilitate management of the cluster using the primary blade. This IP address, which is also known as the cluster management IP address, is always owned by the primary blade. The IP address ensures that when a BIG-IP Administrator connects to the IP address, the administrator is connecting to the primary blade, regardless of which blade is currently the primary blade."
So, the primary blade as per your screenshot would be 2/mgmt at the moment, I believe. You should be able to use TMSH commands, once you SSH into the primary blade to manage the other blade.
Of course, I would also check VLAN and cabling. Make sure the MGMT of each blade is connected to the SAME VLAN/subnet. Also, make sure the switchport is showing the same connectivity for each MGMT interface. It is also strange that you are negotiating to 100/full on the MGMT interface... as I believe those are gig capable and that would be desirable for many reasons.
Hope some of this info helps.