Exchange 2007 – High Availability Solution – Standby Continuous Replication

Standby Continuous Replication is a new feature which is added as one of the high availability method in Exchange 2007 Server. This feature is supported in Exchange 2007 service pack1 only. By the name itself you can understand that standby continuous replication is a standby server which can bought up as soon as possible online so that user can start to access their mailboxes. You implement it as a recovery server in the same datacenter or else you can have a  on another datacenter.

It follows the same log shipping technology as like we have for LCR & CCR. The starting point of log shipping process is called as Source Server and the place where it is log shipped is called Target Server. SCR is supported for single mailbox server,LCR CCR & SCC.
Inorder to enable SCR on the alternate server you need to keep the alternate server in the same Active Directory Domain.

Difference between SCR as compare to LCR and CCR:
 In LCR and CCR we use to have LCR or CCR enable per storage group from source to target storage group on 1:1 ratio i.e one to one copy. However in SCR we can have a copy of storage group from source to many target server on 1:many ratio. But Microsoft always recommend to have maximum of 1:4 ratio.
 LCR and CCR can be enabled from powershell or GUI but SCR is enabled from Powershell command only.
 SCR will have delay in the log shipping technology because if we have logical corruption in the database still administrator can start the standby mailbox server for mail access.
 You cannot take the backup of target copy database.

SCR Deployment:
 SCR can be deployed from one storage group to another storage of another mailbox server which is located in different site (Single Mailbox Server concept between the site)
 

 SCR can be deployed in a CCR enabled server i.e having CCR enabled in one site and SCR enable from the CCR server in the different site.
 

 We can have one source storage group copy to many target storage group copy.

Command:
Already we are aware about all the basic commands for continuous replication which we have done in LCR and CCR, in SCR they have comeup with a new command “Enable-StorageGroupCopy” for enabling the SCR.

Test-ReplicationHealth
Get-StorageGroupCopyStatus -Identity <Server>\<StorageGroup> -StandbyMachine

Fine these are the basic overview..

I would like to show you the standby continuous replication for standalone server

My Lab Setup:
Site1:
                        MainDC – AD\DC\GC\DNS
                        Hub-Cas – holding the role of hub, cas and mailbox server

                        Edge Server – having 2 nic cards for connecting the site1 and site2

Site2:
                        Site-B – AD\DC\GC\DNS
                        DRE2K7 – Having three roles installed i.e hub,cas and mailbox server.

Thumbrule : All the servers for whom we need to configure standby continuous replication i.e Source and Target Server should be in the same domain.

I will show you how to enable SCR for one storage group and mounting the DR database.

Figure 4:
 

In the above figure I have created a storage group called as “SSG” and mailbox store called as “MBX-SSG”.
I have created one test account called as SSG-MBX. And also I have sent few test mail.
Figure 5:
 

Now lets check the size of the MBX-SSG.edb
Figure 6:
 

Configuration of SCR:

Now I need to configure SCR (standby continuous replication for SSG, so that if my SSG group get corrupted or my site-1 get effected I should have another site ready with me for mounting the database of SSG.

Inorder to do this:
Log on to DRE2K7 Server (This is my recovery\standalone server in other word it is my target server)

To enable SCR we need to type the following command from the power shell:
Enable-Storagegroupcopy sourceserver\Storage group -StandbyMachine TargetServer -ReplayLagTime 0:0:0

-ReplayLagTime  : This parameter is used to specify the amount of time that the Microsoft Exchange Replication service should wait before replaying log files that have been copied to the SCR target computer. The format for this parameter is (Days.Hours:Minutes:Seconds).

Note : I have set replay lag time to zero because it is lab test to education. However in realtime the default period for replaylagtime is 24 hours which can extended to 7 days or reduce to 0 seconds. Default delay in log replay activity of 50 log files. You can change the replay lag time however for default delay activity for 50 log files are hard coded you can’t change it. After being set, the value for this parameter cannot be changed without disabling and then enabling SCR.

One more thing I would like to add here is, as I have 2 different site or let say whenever you are configuring SCR between the site make sure that you are Replicaiton is working fine between the site and do a forceful replication whenever you make changes otherwise you will not get the output as you expected.

In my case I will type:
Enable-Storagegroupcopy hub-cas\ssg -StandbyMachine DRe2k7 -ReplayLagTime 0:0:0
Figure7:
 

Now I have enabled SCR for SSG
Next step I would do is force a replication between the sites, for forcing the replication between the sites I used Replmon tool.

Now you can see one folder created under C root drive of DRE2K7 after Enable-Storagegroupcopy
Figure 7.1:
 

However inside the SSG folder you won’t see any log file nor database because we need to manually seed the database and resume it back for log shipping process.

How to do seeding:
Now inorder to seed the database we need to suspend the continuous replication, then seeding the database and resume it back.

For suspending the continuous replication we need to type the following command:

Suspend-StorageGroupCopy sourceserver\Storagegroup -StandbyMachine Target Sever

i.e,
Suspend-StorageGroupCopy hub-cas\SSG -StandbyMachine DRE2K7
Figure 8:
 

Seeding the database : Update-StorageGroupCopy sourceserver\Storagegroup -StandbyMachine Target Sever
i.e.,
Update-StorageGroupCopy hub-cas\SSG -StandbyMachine DRE2K7
Figure 9:
 
Note : After running update command it has seeded the database on a target server.

Here I will run Replmon Tool for AD replication.

Even you can check the status of the storage group copy by running this following command
Get-storagegroupcopystatus sourceserver\SSG –StandbyMachine TargetMachine

Get-storagegroupcopy status hub-cas\ssg –standbymachine DRe2K7

Now after this I need to resume my suspend continuous replication and inorder to do that I need to run following command:
Resume-StorageGroupCopy sourceserver\Storagegroup -StandbyMachine Target Sever
i.e.,
Resume-StorageGroupCopy hub-cas\SSG -StandbyMachine DRE2K7

After resuming the storage group, check the status of continuous replication by running this command
Get-StorageGroupCopyStatus hub-cas\SSG - StandbyMachine DRE2K7 |Fl Summary*,Copy*
Figure 10:
 

In the DR Server you can see the database and log files which are shipped
Figure 11:
 
We have enabled SCR now inorder to mount the SCR database we need to have one Storage created and a mailbox store.
Note: If you create a same mailbox store or storage group name as like we have in production environment it will get conflict so we need to create a different storage
I am creating a storage group called as DRSSG and Mailbox Store called DRMBX and place I am storing the database in C:\DRSSG

Either you can create it from powershell or GUI
From powershell:
New-StorageGroup -Server DRe2k7 -name DRSSG  -LogFolderPath c:\DRSSG -systemfolderpath c:\DRSSG

New-MailboxDatabase -StorageGroup DRe2k7\DRSSG  -Name DRMBX  -EdbFilePath c:\DRSSG\DRMBX.edb

Mount-Database DRMBX

Dismount-Database DRMBX

Del c:\DRSSG\*.*

Note: I told you earlier we don’t require to create same store as like we have in production server so we need to create a new storage then we will do something like dial tone process for redirecting the production database to the recovery server databaase. Even you will notice that I have deleted all the content from the DRSSG folder.

Figure 12:
 

Figure 12.1:
You can see the DRSSG folder is empty:

Now we have successfully configured the SCR between the sites.
Note : while doing configuration you might face some issue whenever you are configuring SCR. I will give you some tips regarding troubleshooting of configuration and some workaround later in this article.

My data is working fine and SCR enabled. Now lets assume my database under SSG which is located in the production environment (Hub-Cas – Server name) got corrupted or may be server goes down due to uncircumstances situation. Now I require to make the database available to my resource. Now let see how we enable it as soon as possible.

Redirecting the data to our reovery Server (DRE2K7)
 As data got corrupted in my production environment, the first thing I need to do is dismount the database. So lets dismount the database :
Dismount-Database SourceServer\Storagegroup\database -Confirm:$False
 Then restore the database this is nothing but making the data avail on the dr server.
Restore-Storagegroupcopy SourceServer\Storagegroup\database -StandbyMachine TargetServer
Restore-Storagegroupcopy hub-cas\SSG\MBX-SSG -StandbyMachine DRE2K7
 Now check the database status of the replicated database on DR Server
Note: It will be always dirty shutdown
eseutil /mh “path of the database” from the bin directory
eseutil /mh “C:\SSG\MBX-SSG.edb”
Figure 13:
 
 Run soft recovery command : eseutil /r e02
Figure 14:
 
 Now again run eseutil /mh command to know the status of the database : This time it should be "Clean Shutdown"
Figure 15:
 
 Now we need to move the storage group path & database path to the newly created storage group- for more clarification on new storage group name see the figure 12 & 12.1(Note : we are just moving the configuration file only)
Command : move-storagegrouppath dre2k7\DRSSG -SystemFolderpath c:\SSG -Logfolderpath c:\SSG -configurationonly -confirm:$False

Move-DatabasePath dre2k7\DRSSG\DRmbx -edbfilepath c:\ssg\mbx-ssg.edb -ConfigurationOnly -Confirm:$False
Figure 16:
 
Note : with the configuration mode only the path are updated not the databases are moved

 we need to allow file restore, so run the following command
Command : Set-MailboxDatabase dre2k7\DRSSG\DRMBX -AllowFileRestore:$true
 now we can mount the DRMBX database
Mount-Database DRMBX
 Once you mount the database, now you can able to understand what does mean by configuration mode movement which I described in Figure 16
Figure 17:
 

Note : the path of DRSSG database right now it is showing C:\SSG but if you notice figure 12 & 12.1 the path was C:\DRSSG.

Now we are done with 95% of task, but what about the users who are located under SSG in the production, will they redirected automatically to DR server. Answer is No.
Figure 18:
 
Note : SSG-MBX is my test user which is still reflecting to my dismounted database of my production server.

 inorder to redirect the mailbox to the DR server, we need to run the following command:
Get-Mailbox -Database hub-cas\SSG\MBX-SSG |where {$_.objectClass -NotMatch '(SystemAttendantMailbox |ExOleDBSystemMailbox)'}| Move-Mailbox -ConfigurationOnly -TargetDatabase DRE2K7\DRSSG\DRMBX -Confirm:$false
Figure 19:
 

 Now lets take a loot at the Mailbox – ssg-mbx (my test mailbox)
Figure 20:
 

Note: Now my mailbox is redirected to DR server.

 Let me check the mails in my test mailbox
Figure 21;
 

This how we can configure for site resilience.

Troubleshooting:
1) Whenever we configure SCR between the sites, major problem you will face is delay in the replication process for changes occurred. So use Replmon tool for replicating. I used it at each and every place eventhough I had good network connectivity.
2) Check the replication status after configuring the SCR like : run 
        a. Test-replicationhealth
        b. Get-storagegroupcopy status
Note: Your copy status should show healthy conditions or else your log file will not be replicated and can’t do restore storage group.
 you can try some the of the step when your storagecopy status is unhealthy:
        o suspend 
        o update
        o resume
        o get-storagegroupcopy
        o test-replication health.
 Whenever your replication status is healthy you will see the shared folder something like this Figure 22: in the production server
 

If you are not able to see that share icon.. you can do small trick here, which I did for my lab server.
 Run test-replicaitonhealth command on Dr server from the powershell window
    o This will give you one guid value : example The directory \\HUB-CAS\9b1c34e2-eda1-4688-8a08-a119a3ad4899$ required by the Microsoft Exchange Replication Service for HUB-CAS\SSG does not exist. Check the file system and its permissions.
    o In the application log you will see the following event:
        Event Type: Error
        Event Source: MSExchangeRepl
        Event Category: Service 
        Event ID: 2074
        Date:  2/21/2008
        Time:  4:41:42 AM
        User:  N/A
        Description:
        The directory \\HUB-CAS\9b1c34e2-eda1-4688-8a08-a119a3ad4899$ required by the Microsoft Exchange Replication Service for HUB-CAS\SSG does not exist. Check the file system and its permissions.
 First thing try to restart the Microsoft Replication Service. And replicate your AD.

If then also not working - small workaround which helped to resolve (small trick which worked for me)
 Then make a note of the guid value which you are able to see while running test-replicaitonhealth
 Right the storage group folder for which you want to enable SCR => Click on Share Tab, then paste that guid value in the share name and in the comment : \\servername\storagegroupname and then add Exchange Server group in Permission & security and give me full control or atleast special permission.
Note: Under Permission : Everyone will be added remove that group
Figure 23:
 

I hope that this article is being very informative for you all. Thanks a lot for spending your time in this matter.
Exchange 2007 high availability – Good improvement by Microsoft.

I would like to give special thanks to Microsoft team, msexchangeteam and msexchange.org website because without them I can’t reach to this level where I can share my view regarding this high availability.

Related Articles:
Exchange 2007 - Local Continuous Replication (High Availability)

Exchange 2007 – High Availability Solution – Cluster Continuous Replication

Exchange 2007 – High Availability Solution - Single Copy Cluster

Reference Article:

Standby Continuous Replication
Standby Continuous Replication: Site Resilience with Standby Clustering