Wrong wireless IP addresses after dcpromo
Last week we were shut down for factory maintenance. We brought a new primary domain controller on line and demoted the old, and moved DHCP to the new PDC. Since then our wireless devices have been picking up the wrong address. For example, wireless devices should pick up an xxx.xxx.133.xxx address, but report they have an xxx.xxx.130.xxx address, and then they do not connect to the network. We did not have an idea of the full impact, but the wireless network is basically down. There are a hand- full of devices that have the proper address. I don’t have a clue where to begin. Normally, I would check the network switches, and I did briefly. However, this started when we switched domain controllers. So, I think it might a better use of time to look at the DHCP side, which I have briefly. Where and what do you think I should start looking at?
Mahesh
check dhcp scopes if they are not reflected with correct dns server IPs after domain controller switch over
ASKER
I forgot to mention we moved DHCP to the new PDC.
When you moved DHCP, how did you do it? Did you export the scopes to a file and then reimport them on the other side.
Were the OS versions the same or not? Did you move from 2003 or 2008 to 2016 or something?
Were the OS versions the same or not? Did you move from 2003 or 2008 to 2016 or something?
Have you checked how the DHCP server on the new DC is configured?
Are the VLANs? If so, is the switch port configured correctly? (Should be already, but worth asking)
Are the VLANs? If so, is the switch port configured correctly? (Should be already, but worth asking)
ASKER
More comments:
Also, one domain controller was originally setup to have to have addresses 1-127 and the other 128-253. This is still the setup.
1. We are very low on addresses in the suspect scope. I relocated 30 some devices to a new scope/VLAN and deleted the old lease. However, I just checked the scope and the scope is back to 100% full.
2. The network switches remained the same.
3. We demoted the old because we wanted the new PDC to have the same address.
4. I had deleted 30 some leases on the computers after I relocated them to a different scope/VLAN.
5. They don’t seem to be defaulting to APIPA address, just the wrong scope. It’s like they get the wrong address for the scope they are on and then can’t connect even though they have an address (though the wrong address).
Also, one domain controller was originally setup to have to have addresses 1-127 and the other 128-253. This is still the setup.
1. We are very low on addresses in the suspect scope. I relocated 30 some devices to a new scope/VLAN and deleted the old lease. However, I just checked the scope and the scope is back to 100% full.
2. The network switches remained the same.
3. We demoted the old because we wanted the new PDC to have the same address.
4. I had deleted 30 some leases on the computers after I relocated them to a different scope/VLAN.
5. They don’t seem to be defaulting to APIPA address, just the wrong scope. It’s like they get the wrong address for the scope they are on and then can’t connect even though they have an address (though the wrong address).
ASKER
Alex, We did import export the file. we migrated from 2003-2016
ASKER
Masnrock, The DHCP appears to be configured correctly and I checked a few related ports and they seem ok.
ASKER
Mahesh, the DHCP reflects the correct DNS IPs.
Right,
I've seen this before, 2003 does a really bad export and you normally have to jetpack the database. Run another export on the 2016 box and then delete the scope on the 2016 box and then re-import it using the 2016 export.
I've had it do silly things before, it's long winded but this may help.
Reagrds
Alex
I've seen this before, 2003 does a really bad export and you normally have to jetpack the database. Run another export on the 2016 box and then delete the scope on the 2016 box and then re-import it using the 2016 export.
I've had it do silly things before, it's long winded but this may help.
Reagrds
Alex
Or you could just redo the DHCP portion manually. Should be far cleaner.
I also should ask whether the DHCP server listed on clients getting addresses was the correct one. If so, it at least keeps the scope of troubleshooting a bit narrower.
I also should ask whether the DHCP server listed on clients getting addresses was the correct one. If so, it at least keeps the scope of troubleshooting a bit narrower.
it won't take the current leases over however, that's the only thing.
ASKER
A TCPIP stack reset helped one individual, but not others, if that means anything .
Alex, Thank you. That sounds promising. Should I delete those leases?
Alex, Thank you. That sounds promising. Should I delete those leases?
ASKER
Masnrock, we have numerous exceptions and reservations. Do you think the export/import would be good enough?
Yes, delete the incorrect leases, you could also force an IP Renew using powershell maybe.
Otherwise, just export and import again on 2016, 2003 sucks.
Otherwise, just export and import again on 2016, 2003 sucks.
ASKER
Alex, the import said that the scope already existed, so I deleted it and tried the import again, put the scope won't come back. my thought is that since the import errors because the other scopes still exist, the import aborts. Should I delete all the scopes? I'm nervous about doing this.
ASKER
okay I deleted all the scopes and re-imported and they came back.
*Ideally* import/export would work fine, but I always suggest to be prepared for cases when things go sideways.
OK are they now presenting correct IP addresses to your network? And yes, you'd need to remove them all if you exported them all. Then it'll do a new import with the correct database so you don't have to jetpack it.
You might wanna restart your DHCP server service too. Just to be nice :-)
Let me know if it works
You might wanna restart your DHCP server service too. Just to be nice :-)
Let me know if it works
ASKER
Alex, this seems to be working
ASKER
don't want to throw my hat in the air yet. don't want to jinks. but I think we owe you.
ASKER
things seem to be going down hill again, only not as bad. most devices got the write address, and now some are going back to the wrong address. I did delete the old address. and restart DHCP. maybe it needs time to stabilize?
ASKER
I noticed the Domain and Forest functional levels were both set to 2003, so I have raised them to 2008R2 since that was an available choice. Waiting to see what happens.
That won't impact your DHCP,
Hows it going today?
Hows it going today?
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.