We help IT Professionals succeed at work.

DCPROMO 2008 R2 OUT - SLOWNESS AFTER FUNCTIONAL LEVEL RAISED TO 2012 R2

hi,

Variety of performance problems after dcpromo out a 2008 r2 dc. General slownes of apps edms - some scripts run slower.
Very hard to pinpoint cause. ran dcdiag all clear - all DNS reference to old dc gone - no legacy mapped drives or any connection to old dc to be found'
raised functional level of domain to 2012 r2  - we have 2 dcs very simple single site domain. now just 2 2012 dcs

the fsmo roles all accross to 2012 r2 dc. - 1 thing that did happen is the fsmo holder server had many patches applied not the wannacry patch that causes slowness.
the point is is that the 2012 r2 dc is just that just a dc with dns - its not used for anything else. i dont see how it could cause slowness accross applications
I have read that SMB signing can cause slowness - when you run 2012 - thing is the 2012 dc/s have been in place for years no issue.

before we dcpromoed out the 2008 rs server we turned it off for 2 weeks - and had no issues. Its only post the dcpromo and 2012 functional level up and perhaps the patching of the dc that the weird sslowness in some apps has raised its head.

what can raising the functional level and dcpromo do to slow the network. ?? thanks
Comment
Watch Question

IT Architect
Distinguished Expert 2017
Commented:
To my knowledge, Raising FFL won't cause slowness issue. You might need to check, load on DC. Can you check handles count for lsass.exe process. If it is above 50K restart the server and check again. Below are few things i might check.

1) How is network utilization for your DC's.
2) Next what protocol your applications are using, NTLM or Kerbrose. Note: NTML is not a good protocol.
3) Any 3rd party tool you are running on DC, like change auditor or tivoli or any monitoring agent, if yes disable them and see, if that makes any difference.
4) Enable netlogon logs and check it.
5) Enable field engineering logs and check, which application adding load on your DC's.

Finally, open MS case, and ask them to take full memory dump and check for the issue.

Author

Commented:
Thanks - it was not related - nessus scan left an interface in inconstant state - reboot of sun box resolved