Hi,
I have small cluster with few host, 3 standard switches (management network, vMotion and software iSCSI) and 1 distributed switch in cluster. Recently I started implementing auto deploy to eliminate boot disks in hypervisors. I've created vlan for auto deploy in my network, added it to distributed switch. Migration of first server to auto deploy was quite easy - I boot it using auto deploy, added it to vDS with assigning port group, then creating configuration, creating host profile and apply host profile. Everything is working ok - host after boot apply it's profile, is part of vDS. This profile works well on next two servers too.
But I have problem with 4th server. After manual apply of host profile is everything ok but after reboot host several things from the host profile is missing. The most several thing is that the host after boot isn't part of the distributed switch - interface with auto deployed IP address I can see under standard switched (see screenshot - vmnic0 should be under distributed switch) and in distributed switch configuration I can see that this host is out of sync. When I put host in maintenance mode and apply profile again it finishes configuration and everything is ok.

I also notice that boot takes very long time (about 20 minutes) - if I detach profile and host boot without configuration it takes max 5 minutes. In auto deploy log I found these lines:
2013-03-08 09:49:46,609 [5032]WARNING:addhost:abaf5218acac8a02755841361fb82ec5:no selected management vnics!
2013-03-08 09:49:46,609 [5032]WARNING:addhost:abaf5218acac8a02755841361fb82ec5:falling back to phone home address
2013-03-08 09:49:46,609 [5032]INFO:addhost:abaf5218acac8a02755841361fb82ec5:selected IPs -- [EsxAddress[{'addr': '10.10.110.117',
'family': 2,
'hostname': 'nms5.man.casablanca.cz',
'origin': 'dhcp',
'status': 'preferred'}]]
2013-03-08 09:49:46,673 [5032]INFO:addhost:abaf5218acac8a02755841361fb82ec5:using current address -- EsxAddress[{'addr': '10.10.110.132',
'family': 2,
'hostname': 'nms5.man.casablanca.cz',
'origin': 'manual',
'status': 'preferred'}]
2013-03-08 09:49:46,703 [5032]INFO:addhost:abaf5218acac8a02755841361fb82ec5:host profile engine found, checking state
2013-03-08 09:49:47,219 [5032]INFO:addhost:abaf5218acac8a02755841361fb82ec5:reapply required for -- (str) [
'DvsProfile'
]
2013-03-08 09:49:47,358 [5032]ERROR:addhost:abaf5218acac8a02755841361fb82ec5:host requires reapply, but is not in maint mode
2013-03-08 09:49:47,375 [5032]INFO:addhost:abaf5218acac8a02755841361fb82ec5:host profile reapply required -- False
This explains why it cannot reapply host profile. But previous three hosts correctly boot in maintenance mode and reapply works fine.
This server has different hardware from previous (this is HP 1U server, previous was custom builds with intel motherboard). I'm using different host profiles for it. First I tried cloned profile from previous servers (with updated network settings) after I try to configure host from zero with only networking, create profile and reboot but results was same.
Do you have any idea what can prevent this server from booting in maintenance mode?