IP Hash Nic Teaming and VMkernel mgmt portgroup

[kwg66]

This is a simple question, with IP Hash configured at the master vSwitch (standard vSwitch), all port groups are supposed to have all VMNICs in the active state, but does this rule also apply to the vmkernel mgmt.  portgroup? 

I had read that vmkernel mgmt. bypasses IP HASH rules, but I would like clarification about this please.   We have had 2 hosts become unresponsive periodically and I had noticed that the vmkernel mgmt. portgroup didn't have both vmnics set to active, only 1 was active and the other was standby.   I want to verify this didn't cause network issue that resulted in our hosts become unresponsive.

Thanks in advance.

[bayupw · Accepted solution]

As per Route Based on IP Hash, when using IP Hash You must set all uplinks from the team in the Active failover list. The Standby and Unused lists must be empty.

On the intermittent side of things, it is probably similar situation described in this KB NIC teaming using EtherChannel leads to intermittent network connectivity in ESXi (1022751)

[kwg66]

Even after I addressed the improperly configured mgmt. port group we still had the hosts become unresponsive, we upgraded the firmware to the latest, so much for Cisco's "preferred and most stable" recommendation, the so called stable and recommended was causing the problem.

However, that being said, I did learn that the IP HASH is required in mgmt. port group in all cases of port channel upstream, the link to that article you shared detailed how you can lose your connection during the set up \ switch over to ether channel made it clear to me.   I decided to follow up on the post I had originally read that stated you didn't need IP HASH in the mgmt. port group and sure enough I had mis interpreted it due to reading through it too quickly.   Thanks.