Quantcast
Channel: Exchange Server 2013 - High Availability and Disaster Recovery forum
Viewing all 1985 articles
Browse latest View live
↧

Unable to recover from Database Switchover

July 9, 2014, 9:51 am
$
0
0
Hi
We are running Exchange 2013 DAC mode enabled and are testing a DR scenario where site A goes down and we need to recover in site B

mailsvr1 is the active mailbox server in site A
mailsvr2 is the passive mailbox server in site B
CASsvr1 is the witness server in site A
CASsvr2 is the alternate witness server in site B

We drop Site A and successfully recover site B by running:
Stop-DatabaseAvailabilityGroup dag1 -MailboxServer mailsvr1 -ConfigurationOnly
Stop-Service clussvc
Restore-DatabaseAvailabilityGroup dag1

We bring Site A back up and try to rejoin it by running:
Start-DatabaseAvailabilityGroup dag1 -MailboxServer mailsvr1 -ConfigurationOnly
start-service clussvc

this step fails as the service is disabled, setting the service to automatic and retrying results in:
The Cluster service cannot be started. An attempt to read configuration data from the Windows registry failed with error ’2′.

we try to start the service with:
net start clusscv -forcequorum
this also fails with Error 1722

running the following commands do not help:
Restore-DatabaseAvailabilityGroup dag1
Set-DatabaseAvailabilityGroup dag1

running the following commands do not help:
Cluster.exe Node mailsvr1 /ForceCleanUp
Start-DatabaseAvailabilityGroup DAG1 –MailboxServer mailsvr1

In EAC the DAG shows mailsvr1 is not operational
If we try to add mailsvr1 in FCM, it says that its already a member of a DAG

the output from Get-DatabaseAvailabilityGroup | fl is correct:
Name                                   : DAG1
Servers                                : {mailsvr2, mailsvr1}
WitnessServer                          : CASsvr1.domain.com
WitnessDirectory                       : C:\DAG1
AlternateWitnessServer                 : CASsvr2.domain.com
AlternateWitnessDirectory              : C:\DAG1
ManualDagNetworkConfiguration          : False
DatacenterActivationMode               : DagOnly
StoppedMailboxServers                  : {}
StartedMailboxServers                  : {mailsvr1.domain.com, mailsvr2.domain.com}

Does anyone know how we can start the cluster service on mailsvr1 and get the node to rejoin?
Thanks

shapi

↧
Search

Exchange 2010 DAG Configuration (Mailbox)

July 10, 2014, 6:12 am
$
0
0
I have one DAG in the environment and I have 2 Mailbox Databases in my DAG. Both databases replicate to each other for failover purposes. The AutoDatabaseMountDial is set to "Good Availability" at this current time. I would like the Mailbox server for Florida attached to Florida Mailbox server as Primary and the Texas Mailbox server attached to the Texas Mailbox Server Secondary. I come in the morning and both Mailboxes are attached to Texas, even though I have Florida as number 1 in the Activation Preference and Texas is number 2.

                              Question,

                              1) Is there a way to failover to Activation Preference 2 and if the activation Preference 1 node is back up to automatically fail back?

                              2) When a failover of the mailbox happens all Outlook users are prompted to close and open outlook. This gets real annoying when you have 500+ users. Is there a way to stop this message and auto sync to the new attached mailbox server?
↧
↧

Exchange DAG Failover Logs

July 10, 2014, 12:25 pm
$
0
0

Hi

I am investigating an issue with a third party mobile solution.

I need to find out the failover history for the past few days of a Exchange Database in a DAG.

Anyone know where i can see this?

If this history is not available, then last fail over time would also help.

Thanks

:-)

↧

DAG configurtion

July 10, 2014, 3:49 am
$
0
0
Can we add a ms exchnage server 2010 to a two differents DAGs? 
↧

Exchange 2013 DAG - Witness Server

June 30, 2014, 3:16 pm
$
0
0

Hi,

we have 4 Exchange 2013 Server, 2 Client Access and 2 Mailbox Servers with DAG configured.

We made some maintenance and rebootet the servers step-by-step.

After we rebootet the Witness Server, the DAG Databases get dismounted and was not available anymore, until the Witness Server was back.

Is this normal, when the Witness Server fails, that all DAG Databases get dismounted?

↧
↧

Restore Mailbox Dumpster Data from Exchange 2010

July 12, 2014, 10:46 am
$
0
0

Hi-

We did a restore from a backup to a recovery storage group.  We exported the mailbox of a particular user to PST file format.

Our question:  does the user's dumpster/recoverable items folder automatically get exported in this scenario?  Is there some additional step or command that has to be performed to ensure that the user's dumpster/recoverable items folder is included in the export?  Our concern is that we are not seeing items the user may have hard deleted or emptied out of their deleted items folder.  We want to see the items that were sitting in the dumpster as of the date of the backup.

Thanks-

John

↧

Exchange 2013 Recover Mailbox Database

January 18, 2014, 7:13 pm
$
0
0
hi Experts,

I am logged in as administrator who is also organization manager and I am trying to recover Mailbox Database
Exchange server name is EXCH2013
Database location is default location and I recovered Mailbox and logs to c:\RDB1 from backup
Here are the commands I ran so far

1.      New-MailboxDatabase -recovery -name RDB -server EXCH2013 -EdbFilePath "C:\RDB1\Mailbox Database 1019791786.edb" -LogFolderPath "C:\RDB1"
2.      eseutil  /r E00 /i /d
3.      Mount-Database RDB
Now I try to run the command Get-Mailbox -Database "Mailbox Database 1019791786" | Restore-Mailbox -RecoveryDatabase RDB to finish recovering I am getting error message as shown in attachment
 

Get-Mailbox -Database "Mailbox Database 1019791786" is returning results.
↧

Moving Exchange 2013 from one Hyper-V cluster to another database fails over

July 15, 2014, 12:37 am
$
0
0

Hi all,

We have 3 Hyper-V cluster With 2 Exchange 2013 server on Cluster A and Cluster B.When we move either of them,all databases fail over to the one that is not moved (so the Exchange we move the database fails).

We then have to manually set active database to correct server.Is this by design or some config that is wrong?

thanks! 

Please mark as helpful if you find my contribution useful or as an answer if it does answer your question. That will encourage me - and others - to take time out to help you. Thank you! Off2work

↧
Search

Exchange 2013 MBX in DAG along with Hyper-V and Failover Cluster

February 4, 2014, 6:18 am
$
0
0

Hi Guys! I've tried to find out an answer of my question or some kind of solution, but with no luck that's why I am writing here. The case is as follows. I have two powerful server boxes and iSCSI storage and I have to design high availability solution, which includes SCOM 2012, SC DPM 2012 and exchange 2013 (two CAS+HUB servers and two MBX servers).

Let me tell you how I plan to do that and you will correct me if proposed solution is wrong.

1. On both hosts - add Hyper-V role.

2. On both hosts - add failover clustering role.

3. Create 2 VMs through failover cluster manager, VMs will be stored on a iSCSI LUN, the first one VM for SCOM 2012 and the second one for SCDPM 2012. Both VMs will be added as failover resource.

4. Create 4 VMs - 2 for CAS+HUB role and 2 for MBX role, VMs will be stored on a iSCSI LUN as well.

5. Create a DAG within the two MBX servers.

In general, that's all. What I wonder is whether I can use failover clustering to acheive High Availability for 2 VMs and at the same time to create DAG between MBX-servers and NLB between CAS-servers?

Excuse me for this question, but I am not proficient in this matter.


↧
↧

Converting to a DR environment

July 15, 2014, 7:22 am
$
0
0

Platform

 - Windows Server 2012 R2

 - Exchange Server 2013

I had built a 2 server environment for 100 users, with 1 DAG. Each server has both MBX and CAS roles on it. In addition there is a primary mailbox database on server A that is replicated to Server B and there is a primary mailbox database on server B that is replicated to Server A.

All good so far, and if I force a failure of one server the primary database for that server gets mounted on the other DAG member. The File Server Witness is a Win2K8 R2 server.

I was labouring under the assumption that with simple round robin DNS, I would be able to have a rudimentary failover between the 2 servers from Outlook. Alas testing has shown this not to be the case. Following some further reading it appears that I may need to move the CAS roles off onto separate servers and create a 'CAS array'. I mistakenly thought CAS arrays were only used for failover of MAPI requests, and that with Exchange RPC over http this was no longer deemed necessary.

Current plan is to add either a single CAS to front the mailbox servers, or to put in two CAS servers again to front the MBX servers.

I'm looking for some advice and guidance as to the best approach, and also any suggested steps to achieve at least some ability for clients to fail over from one server to another.

Kind regards

Mike

↧

Removing a CAS server

July 15, 2014, 1:30 pm
$
0
0

I recently had one of my CAS servers crash in an Exchange 2013 SP1 environment. I've immediately deployed another CAS server on a new VM since it was a critical situation. How can I remove the previous crashed server from the environment totally (It's listed in the servers still).

I tried to do a setup /m:RecoverServer on a new VM with the same server name and IP address of the crashed CAS many times and then do unistallation, although I received different weird errors in eachtime! I guess this method is not gonna to work!

↧

cannot flush Exchange logs

July 15, 2014, 7:20 am
$
0
0

Hello,
I wonder if you can help.
We have Exchange 2013 running on Windows server 2012
The server has been installed and running ok for 6 months.
However we've noticed that the Exchange logs have not been flushed during the backup.
Yosemite server backup is installed. The issue has been logged with Yosemite support but they believe it is an Exchange issue.
I have installed Windows Server backup and run a full backup but that has also not truncated the logs.
The Exchange database and logs are installed on the c: drive.
Any help you could offer would be appreciated.
Many thanks,

Ruaridh

Ruaridh Mackintosh

↧

Dag failed the wrong way??

July 15, 2014, 1:56 pm
$
0
0

Thanks for reading my question. We have a main site and a DR site. I'm running a single exchange server in the main site for ~120 users. I created an exchange server in our DR site, removed the default database from that site and created a DAG failover with the main site primary. I also had a file witness in the main site thinking that if the link between sites failed the two parts at the main site would keep those up, only if the primary server failed would it fail to the DR.

Long story short, we lost the MPLS link due to a fiber cut.  Instead of keeping the primary active, the DR site kicked in and the primary failed to stay online.  We had to scramble to foreceably remove the DAG to get back up and working at the main site. 

I now need to rebuild this in a way that works as I want:  With the primary and file witness in the main site, did I do something wrong to make this fail in what I thought was the improbable way?  Is there a better architecture to accomplish this?

Thanks!
Jim

Jim Tall

↧
↧

How to setup redundant Exchange Server?

July 15, 2014, 8:21 pm
$
0
0

Hi,

We have Exchange Server 2013 Standard Edition running on Windows Server 2008 R2 Standard Edition. I want to setup redundant exchange server for this, so that I can use it, in case of main server failure. 

Can someone please let us know how can I do this?

Thanks, Manoj

↧

Issue using Get-DatabaseAvailabilyGroup -Status Exch 2010

June 25, 2014, 3:03 pm
$
0
0

When trying to use this CMDLET (Get-DatabaseAvailabilyGroup -Status) on Exch 2010, I get a powershell error message:

A server-side administrative operation has failed. The Microsoft Exchange Replication service may not be running on ser
ver mail02.domain.local Specific RPC error message: Error 0x6d9 (There are no more endpoints available from t
he endpoint mapper) from cli_GetDagNetworkConfig
    + CategoryInfo          : NotSpecified: (0:Int32) [Get-DatabaseAvailabilityGroup], ReplayServiceDownException
    + FullyQualifiedErrorId : 749A084,Microsoft.Exchange.Management.SystemConfigurationTasks.GetDatabaseAvailabilityGr
   oup

What I don't understand is this server that it's referencing is the witness server to the DAG and SHOULDN'T have this service installed.  Has anyone run into this issue before, and if so, how did you fix it?  I have not tried rebuilding my DAG thus far.

Thanks for any input.

Thanks, Jeffrey - Infrastructure Administrator

↧
Search

Exchange 2013 - Unbound Stretched Active/Active DAG

July 17, 2014, 7:09 pm
$
0
0

I'm designing an Active/Active DAG utilizing a 3-site scenario, with the third site for FSW.

I intend to have 3 CAS+MBX servers in each location.

Site A & B are separately co-located and Site C is geographically distant site for the FSW

Should the WAN connections of Site A go down (for whatever reason), do the passive copies of Site A's DB's on Site B automatically mount?

What happens with the copies in Site A when this is going on, and what occurs when the links are restored? Is there any manual intervention required here and is DAC/Split-Brain syndrome a concern at all?

What if the client doesn't want Auto-failover, can blockautoactivation be set for the passive DB's to -SiteOnly?

↧

Exchange Performance load balancing

July 19, 2014, 2:49 am
$
0
0

We have a single Exchange Server with 2 mailbox databases, MB01 and MB02.

Each user is split among the 2 databases. and Each database has its own hard drive location, e:\maildatabase\mb01 and f:\maildatabase\mb02. now the 2 drives are at 85% capacity and the server;s 32Gb memory is in 95-98% physical and commit memory.

I want to introduce a 2nd Server with same specs to reduce the load on the 1st, how do I go about this?

1. do I bring up the 2nd exchange server and move some mailbox over? and force users to only look for the 2nd?

2. Do I do DAG but it means only 1 server is active which would not help my cause

Note: I have 400 users and 78% are OWA users.

↧
↧

Removing/ Recovering a CAS server

July 15, 2014, 1:30 pm
$
0
0

I recently had one of my CAS servers crash in an Exchange 2013 SP1 environment. I've immediately deployed another CAS server on a new VM since it was a critical situation. How can I remove the previous crashed server from the environment totally (It's listed in the servers still).

I tried to do a setup /m:RecoverServer on a new VM with the same server name and IP address of the crashed CAS many times and then do unistallation, although I received different weird errors in eachtime! I guess this method is not gonna to work!


↧

Exchange 2013 DAG Database Reconciliation From WAN Fail to Manual Site Switchover

July 23, 2014, 12:01 pm
$
0
0

Hello,

I have a question I can’t seem to find any answers to online with regards to DAG’s and reconciliation of the databases after a WAN failure and manual site switchover. So let’s say you have two sites A and B. With A being the active primary site and B being the secondary passive site. Now say the WAN on the primary site fails but internally Exchange is still up and running just you don’t have any external communication.

Now in that time people will still be emailing each other internally generating new data to that active database. Then it is decided a manual failover to the secondary site which has WAN access still will be done. So now you have that site online and it’s generating new data for the database copy. But now you have two databases which have diverged from one and other. How do you go about reconciling that? Everything online is just about whole site failures but nothing specific to this types of scenario.

So is it a case of just accepting the loss of data in site A after you reseed the database from Site B once WAN is restored to A? Or is it a case of copying out the database transaction logs from site A. And then manually replaying them in site B?

-Nicholas
↧

Router configuration for WNLB and Exchange 2010

July 21, 2014, 10:04 am
$
0
0
Hi there.  I've got 2 new CAS server running with Exchange 2010 setup in WNLB unicast mode (yeh, I know this is not recommended by Microsoft, but without any resource, this is the only solution available for load balancing our CAS server).  The purpose is to replace the current CAS server (old hardware) with the new ones in WNLB configuration to provide redundancy.  Long story short, the new CAS in WNLB is working fine in the internal envrionment, now ready to switch over to replace the current CAS server.  I just have a question on the router configuration.  If we do the NAT configuration on the router to point the public IP (in use by current CAS server) to the new CAS WNLB virtual IP, and then add these new CAS server to the Send Connector list, would this cause any email sending issue, eg: email bounce back?  In the past, I've got 2 new DAG server installed (running with all the Exchange roles), and I tried to replace the CAS server with these new DAG servers.  I've added these new DAG server to then Send Connector, yet certain email send out by the DAG servers got bounced back because the reverse DNS record isn't pointing to the DAG servers (which is within expectation because the public IP is pointing to our current CAS server, DNS record mismatch).  I am just wondering will this issue happenes again on the new CAS after I done the NAT configuration on the router (I am going to re-use the public IP, so not going to modify the public DNS record) and added the new CAS servers to the Send Connector list.  On the Send Connector list, I can't add the WNLB virtual IP, I must add the CAS server to the Send Connector list.  My question is, if an email is handled by one of the CAS server in the WNLB to the external network, will it send the email out through the internal IP of the CAS server or it is send by the WNLB virtual IP?  From what I am seeing, it is using the WNLB virtual IP for internal emails.  Just want to clarify this before I proceed with the cut over.  Thanks.   
↧
Viewing all 1985 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>