We told our customers, that their findings are indeed true and  that these attributes, which are returned via get-mailbox are written only at special events, like at creation time of the mailbox or after a move mailbox has taken place. On a DAG the mailbox database can be owned by any of the possible owners of that database. Changing the possible owner of that database is considered to be a circumstance of making sure, that this database is healthy. At any time the smooth switch to a different owner is much more important than the update of all the properties of the mailboxes on that database. The update of the ServerName property with a mailbox comes with a certain performance penalty. Our product group decided therefore not to take that penalty for Database moves to a different possible owner.

So how to overcome this "by design", if You need the Server name, where the mailbox is currently really located ?

The current proposal for a solution is to use "get-mailboxstatistics <mailboxID> | fl Displayname, ServerName" or "get-mailbox | get-mailboxstatistics |fl Displayname, ServerName". If Displayname is too ambiguous, you may use Identity or PrimarySMTPAddress instead.

Unfortunately this fix  was not carried over into Exchange 2013 and recently we received the request from a highly valued customer, to fix the issue also in Exchange 2013. This fix request at the first attempt was rejected due to mainly high complexity of the fix and to at that moment limited applicability (not many customers have reported the issue). Symptoms and cause from the KB article are also valid for Exchange 2013 (currently at Cumulative Update 3 - see https://www.microsoft.com/en-us/download/details.aspx?id=41175)

We are preparing a public KB-article about that fact, which is planned to be published later this year. Till that moment this Blog post can be used as a reference point. We hereby encourage customers and partners, which encounter the same problem with Exchange 2013 and do not want to ignore that circumstance, to either open a Support Service Request with Microsoft, document this circumstance in the comments section of that post and/or send a mail to appssrv at Microsoft.com (preferably with the support case number).

Requirements for BranchCache

Following is a list of operating systems that support BranchCache content server or BranchCache client computer functionality. To successfully deploy BranchCache in a test lab environment, you must use operating systems that support BranchCache.

General Requirements to Server and Client

Server and Clients must be able to communicate to each other. Clients must be in the same Subnet (otherwise Discovery of Cached Content will not work). All Machines must be able to resolvable in DNS or WINS.

Operating systems for BranchCache client computer functionality

To perform the steps in this guide, you must have three physical or virtual client computers that are running one of the following operating systems:

• Windows® 7 Enterprise
• Windows® 7 Ultimate

Operating systems for BranchCache content server functionality

To perform the steps in this guide, you must have one physical or virtual server computer to be used as a BranchCache content Web server that is running one of the Windows Server® 2008 R2 family of operating systems, with the following exceptions:

• In Windows Server® 2008 R2 Enterprise Core Install with Hyper-V, BranchCache is not supported.
• In Windows Server® 2008 R2 Datacenter Core Install with Hyper-V, BranchCache is not supported.

Necessary Installation and Configuration Steps for Content-Server:

In our Lab Environment the Content Server should deliver the Exchange OAB. In this case the
existing CAS-Server is responsible for the Content we want to get, so the CAS-Server will be our Content-server. As prerequisite for CAS the IIS-Server is in place so we don’t need to change anything at this point.

The first neccassary Step is to install the Branchcache Feature from Roles and Features.

After this is done install the B.I.T.S. feature with IIS Extension Subfeature.

At the Server this should be the needed configuration Steps.

Additionaly to verify the functionality you can use Perfmon to Monitor the Branchcache related traffic information.

 Configure BranchCache performance counters on the content server
 

1. 1.   On CAS with Branchcache installed, click Start, click Search programs and files, and type perfmon. In Search results, in Programs, click perfmon.exe. Windows Performance Monitor opens.
2. 2.   In Monitoring Tools click Performance Monitor to view the Performance Monitor graph. To change the performance monitor graph to report view, click the graph toolbar icon that displays an arrow to reveal the drop-down list, and then click Report.
3. 3.   To add BranchCache counters, click the graph toolbar icon that is a green plus sign (+). The Add Counters dialog box opens. In the left pane, scroll to BranchCache Kernel Mode, and click to expand the list of BranchCache Kernel Mode counters. Click Client Cache Miss Bytes, hold down the Ctrl key, and then click Server Cache Miss Bytes, Hash Bytes, and Projected Server Bytes Without Caching.
4. 4.   Click Add, and then click OK.

  Perfmon Content-Server with working Branchcache

 To reset the Branchcache functionallity and the performance counters on the content server use:

        Netsh branchcache reset

 After the Branchcache reset the Perfmon Counters are reset as well.

 Client computer configuration: 
 Neccessary Installation and Configuration Steps for Client Computers to enable BranchCache distributed cache mode using network shell commands

1.   On the BranchCache client computer that you want to configure, click Start, click Search programs and files, and then type command. In search results, under Programs, right-click Command Prompt, and then click Run as Administrator. The command prompt opens with the elevated privileges that are required to run netsh commands.

2.   Run the following command: netsh branchcache set service mode=DISTRIBUTED

3.   To verify that BranchCache distributed cache mode is correctly configured on the client computer, run the following command: netsh branchcache show status. The BranchCache Service Status is displayed in the command prompt window with the following values: Service Mode: Distributed Caching; Serve peers on battery power: Disabled; and Current Status= Running.

To configure BranchCache performance counters on the Client Computers
 

4.   On Client with Branchcache installed, click Start, click Search programs and files, and type perfmon. In Search results, in Programs, click perfmon.exe. Windows Performance Monitor opens.

5.   In Monitoring Tools click Performance Monitor to view the Performance Monitor graph. To change the performance monitor graph to report view, click the graph toolbar icon that displays an arrow to reveal the drop-down list, and then click Report.

6.   To add BranchCache counters, click the graph toolbar icon that is a green plus sign (+). The Add Counters dialog box opens. In the left pane, scroll to BranchCache, and select all underlying counters.

7.   Click Add, and then click OK.

  First Windows 7 Client got Data from Server                                                                                              Other Windows 7 Clients got Hashes from Server but Data from Cache of First Client

To reset the Branchcache functionallity and the performance counters on the Client machines use:

      netsh branchcache reset

and after that

       netsh branchcache set service mode=DISTRIBUTED

These are the steps to make OAB Download over Branchcache possible.

Multi-Tenant Support

https://technet.microsoft.com/en-us/library/ff923272.aspx 

The hosting solution available for Exchange 2010 SP1 includes most of the features and functionality available in Exchange 2010 SP1 Enterprise deployments, but also includes features and functionality that will allow you to create and manage tenant organizations. Microsoft Exchange Server 2010 SP1 will form part of the suite of multi-tenant capable products that will replace the Hosted Messaging and Collaboration 4.5 solution. 

The account provisioning rollout of new Accounts and new mailboxes works via the new-mailbox cmd:
============================================================================

New-Mailbox -Database "Mailbox Database name" -Name "name" -LinkedDomainController "DCName" -LinkedMasterAccount domain\name -UserPrincipalName name@domain.com linkedCredential:(Get-Credential domain\Administrator) -Organization “tenant or organization name” 

The new-mailbox procedure only covers the creation of new accounts and new mailboxes by assigning the appropriate tenant for the particular AD Account.

The –Organization switch is responsible to match the correct tenant. In many customer environments, migrating to Exchange 2010, there already exist the AD accounts that need to be linked to a new mailbox in the appropriate tenant.

To match an existing AD account to a new mailbox we need the Enable-mailbox cmd. Unfortunately the Enable-mailbox cmd syntax options do not include the –Organization option.

This way we cannot connect an existing AD account to a new mailbox in the appropriate tenant.

Solution:
=======   

If you want to enable an existing AD user for a mailbox at a specific organization, you can still use Enable-Mailbox cmdlet. There is some extra work to stamp a few attributes correctly in the AD user before running the cmdlet (here as a sample domain name.test.microsoft.com):

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

1. msExchCU: CN=Configuration, CN=<tenant name>, CN=ConfigurationUnits,CN=Microsoft Exchange, CN=Services, CN=Configuration, DC=name  DC=test, DC=microsoft, DC=com
2. msExchOURoot: OU=<tenant name>, OU=Microsoft Exchange Hosted Organizations, DC=name  DC=test, DC=microsoft, DC=com
3. userPrincipalName: Usually it is <user name>@<tenant domain name> 

After stamping these attributes the Enable-mailbox cmd automatically links the AD accounts to a mailbox in the appropriate tenant or organization.

Event ID: 1040
Level:    Error
Provider: MSExchangeIS Mailbox Store
Message:  The rule "rulename"" with the sequence number 20 was disabled due to the error -2147467259 that was encountered  while applying the rule.

The problem was absolute sporadic.

So the plan was to configure diagnostic logging and to create an extra trace.
This extra trace can be sent in to Microsoft for further analysis.
The extra trace should run all the time and should be stopped after the problem happens.
The challenge here was how to stop the trace, at the time the problem happens, so that the
log file, would not get overwritten.

Here is what you can do if the operating system on which Exchange 2007 runs on is windows 2008:
The steps are:

1.
Set diagnostic logging (in this case diagnostic logging for private rules)

On the Exchange server open the powershell:
Set-EventLogLevel "MSExchangeIS\9000 Private\Rules" -level high

2.
a) Start tracing:

Go to c:\program files\microsoft\Exchangeserver\bin
From there start the extra.exe

Select a task
Trace control; OK
Select trace file location: Enter the path where the file should be saved
Select trace file name: Enter the file name
Enter max trace file size (MB): Enter 250
Select trace file behavior: Circular logging

Select manual trace tags:
- Trace Types: Select all except performance
- Components to Trace: On the left site click on: Store
- Trace Tags: On the right site select:

tagDisableRules
tagPrivateRulesErrors
tagrulelocks
tagRulesMiscellaneous
tagRulesSync
tagRulesTable
tagTriggerRules
tagvrules

Then check the checkbox "show only enabled components" and the checkbox: “Show only enabled tags”.
If other components or tags are enabled then the ones we selected please uncheck them

Click on start tracing. Then let the GUI stay as it is.

b) Configure to stop the tracing based on event 1040:

Attach to the event 1040 in event viewer:
Go to start, run: eventvwr.exe
In event viewer go to Windows Logs, Application, right click on the event 1040 and click on
"attach task to this event"
Give it a name and a description, click on the next button 2 times

Select "Start a program" 
program: c:\windows\system32\logman.exe
Arguments: stop -ets -n ExchangeDebugTraces
click on next and click on finish

c) check if the trace is running:

Start perfmon.exe and check if the session is running:
Go to start, run: perfmon
Data Collector Sets
Event Trace Sessions
There you should see a session called "ExchangeDebugTraces"
It should be shown as running.

d) After the log file is created

After the event occurred in the event log the trace should be stopped.
In perfmon you should no longer see the "ExchangeDebugTraces" session .
The you can close the GUI from extra.exe

To clear the task from the task scheduler:
Start, Programs, Administrative Tools, Task Scheduler
In the Task Scheduler, go to Task Scheduler Library, Event Viewer Tasks and delete the tasks you created earlier.

You can set the diagnostic logging back to none:
Set-EventLogLevel "MSExchangeIS\9000 Private\Rules" -level lowest

The client is unable to open the winmail.dat file and so the email is useless for him.

The reason for this is, that the Microsoft internal TNEF alias RTF (Rich Text Format) was sent out of the Exchange organization. There are different places and settings which have an influence, which you may need to check:

1. Domain level (Exchange server settings):

a) Exchange 2003

 In the ESM (Exchange System Manager) go to, Global Settings, Internet Message Format, select the domain. Default domain is “*” but please configure an extra entry for the external SMTP domain for example dominodomain.com.

If you open the property of the domain and go to advanced, there is a setting "Exchange richt-text format". You can set it to "Always use", "Never use" and to "Determined by individual user settings".

Please set it to never use

b) For Exchange 2007:

Powershell:

Set-remotedomain domainname tnefenabled $false

In the GUI:

Open the Exchange management console, Organization, Hub Transport, Remote Domains, Domain name:

Please set it to never use.

2. Recipient level

Another place where you can configure RTF is at the recipient level.  

a) AD contact

If you got an active directory based contact you can set it to be a non mapi recipient:

Please uncheck the „Use Mapi rich text format) box

Note: For an AD contact there is one additional setting which may need to be set:

InternetEncoding: 1310720

See KB 924240 for more details.

KB 924240        The email message attached with winmail.dat when you send a mail from Exchange to Lotus Notes through a SMTP connector

https://support.microsoft.com/default.aspx?scid=kb;EN-US;924240 

b) For Exchange 2007:

Open the Exchange management console, Recipient Configuration, Mail Contact:

c) The same is true for a personal contact created in Outlook:

d) Even for a one off addresses this can be set. If you enter the smtp address in the to line for example john.doe@domain.com then do a check names, double click on the email address in the to line,

and you will see the same box as before:

3) Other important Outlook settings:

a) Message Format :

Tools, Options, Mail Format, Message Format (Compose in this message format):

Plain Text, Rich Text, HTML

b) Internet Format:

Tools, Options, Mail Format, Internet Format, Outlook Rich Text options:

When sending Outlook Rich Text messages to Internet recipients, use this format:

-       Convert to Plain Text format

-       Convert to HTML format

-    Convert to Rich Text format

4. Outlook overwriting the Exchange settings:

329454 Outlook 2002 message format overrides Exchange server message format

https://support.microsoft.com/default.aspx?scid=kb;EN-US;329454

HKEY_CURRENT_USER\Software\Microsoft\Office\11.0\Outlook\Options\Mail DWORD:
DisableInetOverride Value: 1

When the registry key is set to a non-zero number, the message format does not override the Internet Message Format that is specified by Exchange.

The key DisableInetOverride still exists in Outlook 2003, 2007 and 2010.

5. NK2 cache

If you have Outlook clients where it sometimes work and sometimes not please delete
the name cache files from the outlook clients (delete username.NK2 file(s))

6. More to read:

821750 How to configure Internet e-mail message formats at the user and the domain levels in Exchange Server 2003

https://support.microsoft.com/default.aspx?scid=kb;EN-US;821750

290809 How e-mail message formats affect Internet e-mail messages in Outlook

https://support.microsoft.com/default.aspx?scid=kb;EN-US;290809

(Restore from backup running hard recovery being interrupted by a corrupted transaction log file)     

In other words – can we extract data out of Transaction log files?

The answer is NO. The content is binary.

If we open transaction log files in notepad we only see few data in clear text like:  

------------

From:
To:
Subject:

------------

Here is a sample to figure out what senders and recipients should be informed about potential data loss in case not all log files could be recovered in a restore scenario.

A good explanation using strings.exe to filter information out of transaction log is available in the familiar blog from Scott Oseychik

 “Rough and Tough” guide to identifying patterns in transaction logs

https://blogs.msdn.com/scottos/archive/2007/07/12/rough-and-tough-guide-to-identifying-patterns-in-ese-transaction-log-files.aspx

###########

1.  Download the "Unix for Win32" utilities from https://downloads.sourceforge.net/unxutils/UnxUtils.zip?modtime=1172730504&big_mirror=0

2.    Extract all files from the UnxUtils\usr\local\wbin subsirectory to C:\UNIX

3.    Download strings.exe from https://www.microsoft.com/technet/sysinternals/Miscellaneous/Strings.mspx, and place strings.exe into C:\UNIX

4.    Make a C:\TMP directory (Unix tools need a Win32 equivalent of /tmp)

5.    Make a directory for all your transaction log files (i.e. D:\customers\test), and place all the logs in this dir

6.    From a cmd prompt, navigate to your C:\UNIX dir

7.    Run the following command:

strings -q -n 16 D:\customers\test\*.log | cut -f3 -d: | sort | uniq -c | sort | tee c:\log-output.wri

What this is doing:

·         Identifies all strings in the logs greater than 16 chars

·         Removes the D:\customers\test\E00xxxx.log: from the output

·         Sorts the output

·         Finds all duplicate records, and retains a count

·         Sorts the final output (ending with the largest # of occurrences)

·         Writes all the output to c:\log-output.wri (use WordPad / write.exe to open; notepad.exe mangles the output)

############

Strings.exe command to filter for the following info out of transaction logs

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

------------

From:
To:
Subject:

------------

1. Create a text file called search.txt with exactly following content :

From:
To:
Subject:

2. Run following command and correct folder path appropriate:

strings.exe -q -n 8 C:\UNIX\Trans_Logfilesfolder\*.log | findstr /i /G:c:\UNIX\search.txt >> C:\UNIX\output.txt

(Prerequisites are step 1 to 6 from the blog above)

It does the following:

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

-Look for data with a string length of 8

-handover to findstr.

-filter for criteria out of search.txt

-export result to output.txt

Example for OUTPUT.txt: (info crossed out to hide customer data)

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

C:\UNIX\Logfiles_HA_RE\E033A952.log: From: XXXXXXX <XXXX@DOMAIN_A.de>

C:\UNIX\Logfiles_HA_RE\E033A952.log: Reply-To: XXXXXXX <XXXX@DOMAIN_B.de

C:\UNIX\Logfiles_HA_RE\E033A952.log: Subject: Re: Testmail Analyzer

C:\UNIX\Logfiles_HA_RE\E033A952.log: To: XXXXXXX <XXXX@DOMAIN_C.de

C:\UNIX\Logfiles_HA_RE\E033A952.log: Subject: Testmail Analyzer

C:\UNIX\Logfiles_HA_RE\E033A952.log: from: < XXXXXXX <XXXX@DOMAIN_D.de >

C:\UNIX\Logfiles_HA_RE\E033A952.log: To: < XXXXXXX <XXXX@DOMAIN_E.de >

C:\UNIX\Logfiles_HA_RE\E033A952.log: X-TBoneOriginalfrom: < XXXXXXX <XXXX@DOMAIN_Fde >

C:\UNIX\Logfiles_HA_RE\E033A952.log: X-TBoneOriginalTo: < XXXXXXX <XXXX@DOMAIN_G.de >

1) the Connector contains an Address Space with an underscore ("_") in it, like in the following example:

2) one part of the address of the final recipient (for example, the "Domain"-part) matches this Address Space.

In this case, the sender of the message will receive a Non Delivery Report from the first Exchange 2007 Hub/Transport-Server and the message will not be routed to the recipient.

The Non Delivery Report contains the following error-information:

 #550 5.4.4 ROUTING.NoNextHop; unable to route

This error is returned despite the fact, that the Scope of the Connector is set to "Entire organization" and the Address Space is visible in the Routing Log Viewer in Exchange 2007.

Please note, that this issue can occur in conjunction with all EDK-Gateways for Exchange 2003 and hence is not limited to the Lotus Notes Connector for Exchange 2003.

You can work around the problem by re-defining the Address Space in question so, that the underscore is eliminated from it and replaced by a wildcard. One possible solution for the example above would be, to replace "abcd_ef" by "abcd*". As soon as this change has been replicated to Exchange 2007, the problem is resolved and messages can be sent from Exchange 2007 to the desired destination.