Assign IP to DAG

Question

Hi Team,

I have some questions regarding DAG. My test enviroment is:

1. Azure VM, Vnet (including 2 subnet and one NATgateway).
2. One DC (Windows server 2019 datacenter) in 10.0.1.0/24, the first EX server (Windows server 2019 datacenter+Exchange 2019 CU12) in 10.0.1.0/24, one file witness server (Windows server 2019 datacenter) in 10.0.1.0/24.
3. Another EX server (Windows server 2019 datacenter, Exchange 2019 CU12) in 10.0.2.0.

I tried to create a DAG and assigned 2 ips (for example:10.0.1.235, 10.0.2.235) to it since EXs are in different subnets . But when I tried to add the first server:

=========================================

=========================================

There's no error if I remove these ips from DAG and add servers. It also works if I create a new DAG ( no IP assigned) and add servers. I can then create database copies. However, I will always get above same 0x57 error if create DAG with IP assigned to it. Seems I can only use IPless DAG.

I tried to manually add one cluster (name it DAG4) and add 2 nodes in Failover Cluster Manager. Then I can create DAG4, add ips to it and add 2 servers. In Computer OU, the CNO is also generated and is in active status. In DNS Server, one A record is generated to point to Exchange server 1 and one point to server 2. I disabled this computer object but managing DAG in Exchange requires the object to be in active? (Official document says it should be inactive ?) Activating it then it tells me the there's permission error....... I think this way isn't workable and give up.

I also tried to pre-stage CNO: https://learn.microsoft.com/en-us/exchange/high-availability/manage-ha/pre-stage-dag-cnos?view=exchserver-2019. Still the same 0x57 error

Regarding the official artcles, IP addresses seem to be just an optional configuration?

https://learn.microsoft.com/en-us/exchange/high-availability/plan-ha?view=exchserver-2019#dag-name-and-ip-address-requirements

https://learn.microsoft.com/en-us/exchange/high-availability/deploy-ha?view=exchserver-2019

=========================================

=========================================

But in this article seems I must assign IP to DAG?

https://learn.microsoft.com/en-us/exchange/high-availability/manage-ha/create-dags?view=exchserver-2019

=========================================

In addition to providing a name for the DAG, you must also assign one or more IP addresses (either IPv4 or both IPv4 and IPv6) to the DAG, unless you're creating a Windows Server 2012 R2 DAG without an administrative access point and you aren't assigning any IP addresses to the DAG.

=========================================

The example deployment in above articles tell me that assigning IPs to DAG should also work.

Sorry I'm not very familier in this area, but I really tried many times to no avail. The only solution I can find is to create DAG with no ip configured to it. I want to figure out if I really need to assign IPs to DAG, if yes, then how can I make it work?

Many thanks if anyone can share ideas.

Best Regards

Etta

Answer 1

Unless you have a technical reason to create an IP for the DAG cluster, I wouldnt do it, as there is no reason to do so.

The reason you may need one is if your backup software,for example, required it - otherwise, keep it simple and do not create one!

Answer 2

I switched to VMWare and found that I need to pro-stage CNO according to the article: https://learn.microsoft.com/en-us/exchange/high-availability/manage-ha/pre-stage-dag-cnos?view=exchserver-2019.

1. Create a computer object (CNO).
2. Assign full control of the computer account (CNO) to the Exchange Trusted Subsystem. Then disable CNO.
3. Add CNO Object to Exchange Trusted Subsystem.
4. Wait for CNO, Permissions and Group membership sync to all AD Domain Servers in the forest.
5. On File Server, add Exchange Trusted Subsystem group to local admin.