DCPROMO timeouts

[gurner]

Im pretty sure of the cause of this issue, and have a kind of idea on what will fix it, but i can't think where.

And the only forthcoming MS suggestion in is based on the premise that the server is already a DC

Theres a long story about a DC being formatted by a manufacturer (another thread, another day) and i'm trying to pickup the pieces, and Dcpromo this server.

However it is going to be a 2nd server in a domain, when i start the promo everything starts ok, and it replicates several thousand objects, then times out and stops with the following errors

'the remote procedure call was cancelled'

the Event log makes the following log

'DS RPC Client Event ID 1232'

Active Directory attempted to perform a remote procedure call (RPC) to the following server. The call timed out and was cancelled.

Server:
DC1.domain.co.local
Call Timeout (Mins):
5
Thread ID:
530

Additional Data
Internal ID:
0

--------

I understand from a MS posting that RPC timeout has shrunk from 45mins in Win2k to 5mins in Win2k3?

the suggestion is to increase bandwidth or increase the timeout by adding a key to HKLM>System>CCS>Services>NTDS

Problem is there isn't an NTDS key in services as it hasn't been promo'd yet.

and it is on a GBit backbone with NIC

any ideas?

Cheers

Gurner

http://www.divineparadox.co.uk

 

[aftertaf]

is the original DC a 2003 server?



seems there is a replication problem, have other DCs been part of the domain and since been removed? Check and make sure in sites and services you dont have any odd objects hanging around.

Also, a NTDSUTIL metadata cleanup could help, check

http://www.petro.co.il

for a good procedure on this.

not DNS etc, cos it contacts, replicates some objects, then blocks...
for me, I don't KNOW what the answer is here, but i'd put a shiny Euro on there being some bad data in AD, or in the info on the domain controllers supposed to be present.

in this case, seeing as it has crashed during the process, there could be some residual info from this failed DC attempt in DNS/Active directory (hence NTDSUTIL again...)

If the original DC is 2003, could you try to get a system state backup and run DCPromo using this info, that way there wont be as much replication..... ?
and try your RPC tweak on the existing DC maybe?

best of luck mate, let us know what happens.

Aftertaf
________
Regain control of your PC, at

http://www.debian.org

 

[gurner]

cheers, will do.

it was an original 2nd DC that went pop under overheating, and part of the contract (and manufacturer practises) we sort the OS and they do the Hardware, and we got a call from the end user saying the server had been formatted by the Manufacturer.

it has been bought back up onto the LAN by us under the same name, but there is no trace of AD having ever existed before on it.

it was removed from AD on the other DC as being 'permanently offline... e.t.c.'

but i see no reason why there could still be lots of guff lying around on the other DC, ha

Cheers

Gurner

http://www.divineparadox.co.uk

 

[gurner]

cheers, a quick cleanup of Sites and Services sorted it (fortunately only the local one, there are 50 servers spread all over)

Gurner

http://www.divineparadox.co.uk

 

[aftertaf]

hehe
sounded like something like that...


Aftertaf
________
Regain control of your PC, at

http://www.debian.org