When I try to add the nodes to the DAG I got the error

“You must provide an value for this property”

Decided I better do some reading on TechNet regarding 2010 DAGs and Server 2012 and it turns out that you must pre-stage the CNO (Cluster Named Object)

The documentation states

Pre-staging the CNO is required for Windows Server 2012 DAG members due to permissions changes in Windows Server 2012 for computer objects.

There are 2 methods to achieve the pre-staging.

1. Add the “Exchange Trusted Subsystem” group with Full Control to the CNO
2. Add the first DAG member with Full Control to the CNO

For me I choose to add the “Exchange Trusted Subsystem”

Adding the Nodes to the DAG after making the necessary changes worked without issue

Got the Exchange server bits installed and then tried creating the DAG which failed.

After a bit of analysis it transpires that the customer built the base OS with Windows 2012 Standard and not DataCenter (Server 2012 doesn’t include an Enterprise Edition any more)

Luckily we can convert between editions of Windows Server on the fly via DISM (Deployment Image Service & Management)

TechNet documentation on converting between server 2012 versions can be found here

To determine the current edition of Windows Server installed run the following

• DISM /online /Get-CurrentEdition

To determine which versions the server can be upgraded to run

• DISM /online /Get-TargetEditions

As can be seen from the above screenshot this version of Windows Server can be upgraded to Target Edition : ServerDatacenter

To complete the conversion between editions we use the DISM command with the following format

• DISM /online /Set-Edition:<Version> /ProductKey:XXXXX-XXXXX-XXXXX-XXXXX-XXXXX /AcceptEULA

In the above example the target edition is ServerDatacenter

A reboot of the server and a check of the server edition reveals that is has indeed been successfully upgraded to DataCenter

A public folder replica has been added to the Exchange 2007 servers to include the new 2010 environment.

After verification of data replication all 2007 databases were updated to have their public folder database point to the 2010 servers and the 2007 server removed from the replica in lieu of completely removing the 2007 public folders.

After a few weeks of normal usage users started reporting that they could no longer access their public folders.

To cut a long story short it was decided to add a replica back into Exchange 2007, as the Public Database was now removed I tried to recreate a new one.

Exchange 2007 is running as a Single Copy Cluster and requires cluster dependencies for normal operation.

As 2010 has been introduced into the environment Exchange 2007 throws an error upon creating the Database

After checking the EMC the database has actually been created but at this point it fails to mount due to the missing cluster resource records

To get the database to successfully mount we need to revert to the Cluster Administrator and create a new resource

Open the Cluster Administrator and Ctrl+N to open the new resource wizard

Next select the possible owners

And finally the dependencies. I chose Exchange Information Store and the Logs and Database drives

I also change the Cluster Resource so that if it fails to come online it doesn’t affect the group which would result in a cluster failover

This is only half the story as at this point the database will still not mount.

The next thing that needs to happen is that the Storage Group & Database GUID need to be retrieved. This can be done using the Exchange Shell

We need to register these GUIDs against the private properties of the cluster resource

To do this we need to drop to the command prompt

Cluster.exe res “Public Folders” displays the properties of the resource

Cluster.exe res “Public Folders” displays the private properties of the resource which is what we are interested in.

What needs to happen now is that the private resources needs to be assigned to the Cluster Resource

In both of the above screenshots the network name has been blanked out

We are now finally in a position to mount the public folder and the Cluster is fully operational again.

This process will work for any database not just public folders.

The admin display version from a pre and post CU1 install are shown below. (Ex2K13-01 running RTM & Ex2K13-02 running CU1)

CU1 also brings new Schema, AD & Domain updates so all three need to be prepared before running setup.

To prepare the schema run

• setup.exe /PrepareSchema

To prepare AD run

• setup.exe /PrepareAD

To prepare Domain its

• setup.exe /PrepareDomain

Finally to install the update onto a RTM version of 2013 run

• setup.exe /m:Upgrade

There are only a few click to navigate though before installation begins

Installation took a good 40 minutes on my machines.

If you have been hanging back on the installation of 2013 until CU1 there is no need to install the RTM version first.
As the updates are cumulative run setup straight from CU1 media and your done.

Happy patching.

Exchange 2013 CU1 can be downloaded here

This time around we will cover the installation of the Mailbox role.

Lets get started by covering off the pre-requisites the Mailbox Role

• Server Media Foundation roles
  

  To install Server-Media-Foundation open PowerShell and run the following commands

  Import-Module ServerManager

  Add-WindowsFeature Server-Media-Foundation

  Shutdown –r -t 00

  

• When the server comes back up after its reboot install Microsoft Unified Communications API 4.0 Runtime

   

• Then we install the Office 2010 Filter Packs Version 2.0
  
• And finally its Office 2010 Filter Packs 2.0 Service Pack 1
  

One final reboot is required before we can install the Exchange 2013 Mailbox role.

Once the server is back up we can go ahead and launch the Exchange 2013 setup routine.

The following screenshots will guide you though the install

Once the installation is complete we can check that the Exchange Admin Center comes up as expected.

Open your web browser and enter the IP\DNS name of your CAS server. For me its

https://cas01/ecp

As I installed a trusted certificate yet I will get a security error once I connect. This can be ignored as its expected behaviour

Once you have accepted the warning you will be presented with the Exchange Admin Center

Repeat the installation process for MBX02 and once complete that concludes part 2 of this blog post.

Next time out I will configure up the Client Access Server Client URLs and setup the Kemp Virtual Load Master

The Exchange environment will be split out over 4 servers (2x Client Access & 2x Mailbox)

The backend mailbox will be configured in a DAG and the front end will be load balanced using a Kemp Virtual Load Master 

All servers are running Windows Server 2012 RTM in vSphere 5.1

The lab is setup as per the table below

DC01	192.168.0.1
CAS01	192.168.0.2
CAS02	192.168.0.3
MBX01	192.168.0.4
MBX02	192.168.0.5
Kemp VLM	192.168.0.6

Ill will not be covering the initial OS configuration and all servers have been joined to the domain

In part 1 here we will deal with installation of the Client Access Server. (Configuration guides will come in later posts)

So with that in mind lets start the installation of Exchange Server 2013.

The Client access server requires some pre requisites before we can launch Exchange setup

These prereqs are the Windows Server Foundation feature and Microsoft Unified Communication API 4.

To install the Windows Server Foundation open PowerShell and run the following commands  

Import-Module ServerManager

Add-WindowsFeature Server-Media-Foundation

Shutdown –r –t 00

When the server comes back up after its reboot install Microsoft Unified Communications API 4.0 Runtime

With all the pre requisites now installed we can finally launch the Exchange Setup routine.

The following screenshots walk though the installation.

And that’s part one finished. This process is repeated for the installation of CAS02

In part 2 I will be covering the installation of the Mailbox Server

The follow error was displayed

To resolve this issue I downloaded the Exchange Management Console Troubleshooter Tool EMTShooter

After running the tool it found an error and displayed the following

After reading through the possible solutions I discounted options 2 & 3 as these had already been investigated and the settings were correct.

So with that I settled on option 1.

After examining the GlobalModules of C:\Windows\System32\Inetsrv\config\ApplicationHost.config I could see that the entry for WSMan was indeed missing

I added the entry and closed the file

Refresh the EMC and viola, good as new.

This issue only occurred when calling from PSTN\Mobile across SIP to Lync.

An SBC was in play to convert from SIP UDP to TCP before sending to Lync. The SBC in question was a Disalogic BordNet 500.

After a lot of packet capture and analysis the issue turned out to be the SBC’s B2BUA being configured to Receive PRACK.

With PRACK enabled no Ringback was heard, after disabling the setting Ringback was heard by the caller.

#Begin Script
$MainDBServer = Read-Host “Enter Mailbox Server where Databases are located”
$ReplicaServer = Read-Host “Enter Mailbox Server where existing DBs are to be replicated to”
$DBs = Get-MailboxDatabase -Server $MainDBServer
foreach ($DB in $DBs) {Add-MailboxDatabaseCopy $DB -MailboxServer $ReplicaServer -ActivationPreference “2″}
#End Script

Almost immediately after getting my system up and running I experienced intermittent freezing and a gradual locking up of the system. If audio was playing its would go into a skipping state, repeating the same words over and over again.

At no point did the PC every crash to the “Your PC ran into a problem that it couldn’t handle” screen

After some Googling it turns out that a large number of users are experiencing similar issues.

See here and here for forum posts on the issue.

Various user report that the issue is related to Intel hardware. Numerous solutions have been posed regarding updating the chipset drivers, creating customer power plans, disabling  C1, C3, C6 sleep states in the BIOS, updating graphics drivers.

None of the proposed solution’s worked for me although I do have a stable system now for a few days with no lockups

To resolve my issues I ran the Intel Driver Update Utility which reported Graphics Driver and Wired Networking as having unknown\unsupported drivers installed.

My graphics card is an aging ATI Radeon x1950 and the drivers for this have already been updated\replaced etc. without system stability occurring.

I found an update for the Intel(R) 82579V Gigabit Network Connection and ever since updating the drivers here I have been freeze free and my system at last appears to be stable.

Don’t know if this solution will work for others but its worth a try.

**Update**

The system is still freezing despite the NIC drivers being updated. It was around 3 days before it froze on me and it in general its a lot more stable with the updated drivers.

Another blog post has suggested that installing the Hyper-V role leads to a stable system.

I have installed the Hyper-V role and have not experienced a lockup yet, time will tell.

**Update 2**

Its now been a week since the Hyper-V workaround was implemented and the system has been rock solid.

It appears that this is a valid workaround until Intel\MS get to the bottom of  the real issue