>  you can turn on Change Notification to get quicker replication.

I enabled  Change Notification 10 years ago, it was a wonderful thing to see 
convergence pull in like that. It was also worth tweaking intra-site back then 
but no much anymore.[1]

Recommended Reading for those who are interested

How the Active Directory Replication Model Works - Domain Controller 
Notification of Changes
https://technet.microsoft.com/en-us/library/Cc772726(v=WS.10).aspx#w2k3tr_repup_how_wrys

Enable Change Notification on a Site Link
https://technet.microsoft.com/en-us/library/bb727062.aspx#E0PC0AA

Keep in mind one of the elements of the “you are not smarter than the KCC” 
paradigm is that even if you enable change notification globally, manually 
created connections are not covered. This bites people frequently.
http://blogs.technet.com/b/askds/archive/2013/01/21/configuring-change-notification-on-a-manually-created-replication-partner.aspx

[1] Intra-site notification was 5 mins with 30 sec pause in W2K and changed to 
15 sec / 3 sec in W2K3, I don’t recall seeing that being changed subsequently. 
Brian probably knows
https://support.microsoft.com/en-us/kb/214678


Cliff’s notes for urgent replication
Certain important events trigger replication immediately, overriding existing 
change notification. Urgent replication is implemented immediately by using 
RPC/IP to notify replication partners that changes have occurred on a source 
domain controller. Urgent replication uses regular change notification between 
destination and source domain controller pairs that otherwise use change 
notification, but notification is sent immediately in response to urgent events 
instead of waiting the default period of 15 seconds.

     *   Assigning an account lockout, which a domain controller performs to 
prohibit a user from logging on after a certain number of failed attempts.
     *   Changing a Local Security Authority (LSA) secret, which is a secure 
form in which private data is stored by the LSA (for example, the password for 
a trust relationship).
     *   Changing the password on a domain controller computer account.
     *   Changing the relative identifier (known as a “RID”) master role owner, 
which is the single domain controller in a domain that assigns relative 
identifiers to all domain controllers in that domain.
     *   Changing the account lockout policy.
     *   Changing the domain password policy.


From: [email protected] [mailto:[email protected]] On 
Behalf Of Michael B. Smith
Sent: Friday, September 11, 2015 1:08 PM
To: [email protected]
Subject: RE: [NTSysADM] DCs as DHCP Clients

It’s called “urgent replication” not “emergency replication”.

And you can turn on Change Notification to get quicker replication.

Here is a source:

http://blogs.technet.com/b/kenstcyr/archive/2008/07/05/understanding-urgent-replication.aspx<https://urldefense.proofpoint.com/v2/url?u=http-3A__blogs.technet.com_b_kenstcyr_archive_2008_07_05_understanding-2Durgent-2Dreplication.aspx&d=BQMGaQ&c=hLS_V_MyRCwXDjNCFvC1XhVzdhW2dOtrP9xQj43rEYI&r=TA_mjBT8bS0r8rLrnubGjA&m=btAuFoggS5HpkSj2DFaEnNR8lA2ELW7HImbbZ9tZ_tY&s=Bi3I0hJBgT58S7XZn5vAKU-HHf55HnKCGsch3dcerOc&e=>


From: [email protected]<mailto:[email protected]> 
[mailto:[email protected]] On Behalf Of Michael B. Smith
Sent: Friday, September 11, 2015 3:51 PM
To: [email protected]<mailto:[email protected]>
Subject: RE: [NTSysADM] DCs as DHCP Clients

Brian Desmond is on this list, so he can say with authority…

But as I remember, within a site, replication is very quick. Outside of a site, 
except in cases of “emergency replication” (e.g., password changes) the minimum 
is still 15 minutes.

There are no issues other than the IP address change for DCs. That affects DNS 
and indirectly, clients trying to locate DCs.

From: [email protected]<mailto:[email protected]> 
[mailto:[email protected]] On Behalf Of Charles F Sullivan
Sent: Friday, September 11, 2015 3:36 PM
To: [email protected]<mailto:[email protected]>
Subject: RE: [NTSysADM] DCs as DHCP Clients

Interesting. So even in Azure that’s the default.

I just set up the first DC at AWS a few minutes ago and got the warning that 
the non-static configuration could cause problems for DNS, but I suppose that 
could be for the obvious problem of the IP address changing.

It isn’t that I’m worried about the IP address changing, since there is a 
reservation in DHCP for the DCs, but I thought it may cause problems other than 
an IP address change.

Not to hijack my own thread, but while setting this up, I found that the lowest 
inter-site replication interval is still 15 minutes. These are Windows 2012 R2 
DCs in domain/forest 2012 R2 functional mode. For some reason I thought it was 
possible to lower that to 5 minutes now.

Thanks.

From: [email protected]<mailto:[email protected]> 
[mailto:[email protected]<mailto:[email protected]>] 
On Behalf Of Michael B. Smith
Sent: Friday, September 11, 2015 2:38 PM
To: [email protected]<mailto:[email protected]>
Subject: RE: [NTSysADM] DCs as DHCP Clients

I can’t speak to AWS but I do it in Azure. In Azure an IP address isn’t ever 
released/reassigned until you “Force Stop” a VM which causes the IP, memory and 
vProcs to be deallocated from the VM.

(Azure also has other options – this is just the default.)

From: [email protected]<mailto:[email protected]> 
[mailto:[email protected]] On Behalf Of Charles F Sullivan
Sent: Friday, September 11, 2015 2:33 PM
To: [email protected]<mailto:[email protected]>
Subject: [NTSysADM] DCs as DHCP Clients

Has anyone had to run domain controllers as DHCP clients?

Someone from another one of our IT groups has provisioned some servers at AWS 
for Citrix and the proposal is to add a couple of DCs there. He says “All AWS 
instance should always be DHCP clients”. I think of this as a bad practice, but 
I would think that if it’s the standard at AWS, then lots of others are doing 
the same. So even better would be if I could hear from someone who does have 
DCs at AWS.

I have cloned DCs in an isolated test network, which we regularly use for 
testing. I’ll be connecting those to the DCs I’m building at AWS for testing 
before even attempting this in production. Even if I have no problems in 
testing, I am leery to do this in prod.

Charlie Sullivan
Sr. Windows Systems Administrator

Reply via email to