Deployment fails with timeout as VM takes too long to power on

Products

• VMware Cloud Director

Issue/Introduction

Power on of recently deployed VMs take too long leading to instantiation failure as the automation tool times out before reaching VMs

Environment

VMware Cloud Director 10.4

NSX-T

Cause

/opt/vmware/vcloud-director/logs/vcloud-container-debug.log:

XXXX | DEBUG    | Backend-activity-pool-110149 | VC20VirtualServer              | Unable to find dvpgs [vcId=XXXXX, moref=dvportgroup-XXX] on host [vcId=XXXX, moref=host-XXX] after 4 of 600 retry attempts. |
(...)
XXXX | INFO     | Backend-activity-pool-109942 | NetworkBackingToSwitchMapper   | DvPortgroup [vcId=XXXX, moref=dvportgroup-XXXX] is backing for the network. Adding the dvSwitch [vcId=XXXX, moref=dvs-XXXX] for this portgroup as a requirement.

From vcenter if we search for this network, it is possible to confirm that:

• 2 networks are returned (Opaque Network and Distributed Port Group)
• ESXi hosts for this cluster only have visibility to Opaque when they should see both
• Transport zone in NSX-T is being used for clusters with N-VDS and VDS

Resolution

Contact networking team and review the NSX-T configuration

Hosts in Cluster with N-VDS should be prepared in NSX with an isolated transport zone

If the same TZ is being used with mixed NVDS and VDS the segments created will create in vCenter, Opaque networks outside of the switch and Portgroups inside the N-VDS