Start a Conversation

Unsolved

This post is more than 5 years old

P

1610

March 27th, 2012 13:00

Restore of an individual content database

Hi,

suppose you have a farm made of:

3 Front End systems

1 sql server system

In order to restore an individual content database, I know you have to restore the sharepoint writer and the sql server elements related to the content db  ( Nmm application guide pag.226 ).

But  you have to restore the sharepoint writer ( I mean , the sharepoint saveset ) ON ALL THE FE systems or a single FE is enough?

Apolologizing for the trivial question :-(

Thanks in advance

PierPa

544 Posts

March 27th, 2012 14:00

Hi PierPa,

SharePoint VSS Writer must be registered on at least one web front-end server within a server farm for SharePoint 2007, While for SharePoint 2010, SharePoint VSS Writer must be registered on all web front-end servers.

Yes, you can use one of those web front-end server to perform the individual content database recovery.

Hope this helps you.

Thanks,

Ahmed Bahaa

1 Rookie

 • 

77 Posts

March 27th, 2012 14:00

Ahmed,

suppose you have a backup group with all the three FE systems in backup: if one of them fails for whatever reason, you have two other copies of the same objects, isn't it ?

But you mean that is enough to use only one of the three FE system for the individual content database restore. I'm right?

Thanks

PierPa

544 Posts

March 27th, 2012 14:00

Hi PierPa,

For distributed configurations, register the SharePoint writer on at least one of the web front-end servers. This web front-end server will be included in the NetWorker backup schedule, and will be used to perform recoveries. So the one front-end you used for backing up is enough to be used for recovering.

Hope that this helps you.

Thanks,

Ahmed Bahaa

1.7K Posts

March 27th, 2012 22:00

Hi again PierPa,

Forgot to mention that, if what you need is to restore some information within the Content DB you can also use PowerControl Kroll Ontrack (third party software) to perform granular restores of SharePoint, by restoring locally the SQL content DB and then with Kroll On Track you can browse the content and perform granular restores.

You can find more information in Powerlink:

https://powerlink.emc.com/nsepn/webapps/btg548664833igtcuup4826/km/live1//en_US/Offering_Technical/Technical_Documentation/300-012-860.pdf

Thank you.

Carlos.

1.7K Posts

March 27th, 2012 22:00

Hi PierPa,

If you need to restore only a content DB you can do it anytime with no impact on the SharePoint environment, and no need to restore any SP writer.

You can restore locally a SQL DB or perform a directed SQL restore of a single DB. That won't have any impact on production, while restoring the SP Writer can lead to a data lose, as you will be performing a PIT restore, so any new data added after that PIT backup will be lost.

Please consider that when planning the restore, and engage you SP admin to ensure you are not overwriting any data.

Thank you.

Carlos.

1 Rookie

 • 

77 Posts

April 3rd, 2012 01:00

Carlos,

the documentation, for a novice like me, is confused/confiusing.

To restore an individual content database it says to restore the FE information related with the individual content db ( Nmm application guide pag.226 ) and ALSO the individual content db from the SQL server. If the documentation is wrong, please correct it. If I missed something, let me know...

Thanks for your patience

PierPa

1.7K Posts

April 3rd, 2012 03:00

Hi PierPa,

Maybe the issue is with the terminology, I don't know if you are SharePoint admin or your knowledge of SharePoint, but please take a look at the following link which I think will help to have a better understanding:

http://msdn.microsoft.com/en-us/library/cc264326%28v=office.12%29.aspx

http://msdn.microsoft.com/en-us/library/cc264010%28v=office.12%29.aspx

Anyway please note the difference between restoring a single DB and restoring a WEB Application.

You should be able to restore a SQL DB as per the documentation. What you do afterwards with that DB is a different story.

Basically the aim of the procedure described in page 227 of the App Guide is to restore not only the SQL content DB, but also the related SharePoint components, so when you restore the SharePoint farm/web application is not inconsistent. If you restore a SQL alone, and then let's say you replace it within the farm, there sill be most likely an inconsistency in the farm, as the DB has a PIT which is not the current time, so any transaction or new data added to the SQL is missing.

Let me know if this helped to clarify or if you need any further information.

Thank you.

Carlos.

1.7K Posts

April 3rd, 2012 05:00

Hi PierPa,

Documentation is correct, I've tested that myself, however would be good to have the SharePoint admin around so that he can check the procedure.

Again, if you just want to restore the SQL to copy it somewhere and then the SP admin will play around with it, or do whatever he/she needs to do, then I would suggest not to follow that procedure, as in this case we would be talking about a directed restore of a SQL DB, and in this case there would be no impact whatsoever on SP environment.

Now, if starting the restore from WFE server, basically it will initiate a SharePoint writer restore, then it will detect the dependency of the SP writer with the SQL Content DB, so then you will need to connect to the SQL DB, open NMM GUI, restore the Content DB and then, once restored you go back to the WFE server and click in restore to finish up the restore.

This is only to keep tracking and maintain consistency in the SharePoint environment after a SQL content DB restore, as there is dependency between SP writers and services, SQL DB's, IIS and SYSTEM COMPONENT savesets, as SYSTEM COMPONENTS also backs up the IIS writers and files.

I hope to be clear enough, but in other words, YES, that procedure is correct, I just wanted to let you know the possible consequences or only the difference in restoring a SQL DB alone or do it starting it from WFE server.

By the way, I cannot change the documentation as I'm not an EMC employee

Thank you.

Carlos.

1 Rookie

 • 

77 Posts

April 3rd, 2012 05:00

Carlos,

thanks a lot.

PierPa

1 Rookie

 • 

77 Posts

April 3rd, 2012 05:00

Carlos,

let me point out that I'm NOT a sharepoint user/admin, but only a simple networker admin, unfortunately

I've never had so many difficulties in implementing backup with modules as I have with NMM and Sharepoint.

Since I'm a dumb user, I trust totally in the documentation.

If I have to restore a individual content database, I follow the documentation line by line, the one related with Sharepoint 2010, and it says:

Performing SharePoint Server 2010 individual content database recovery

 

Perform the following steps to recover documents or items that were corrupted:

1. Open the NMM client software on the web front-end where the SharePoint

services writer was registered at the time of backup, and start recovery of

SharePoint writer.

2. Browse the SharePoint save sets. ( HERE I CHOOSE MY CONTENT DB under APPLICATIONS\Microsoft Office Sharepoint Services)

3. Click Recover

A dependency dialog box appears, alerting you to go to the specified remote SQL

Server host and start the recovery processes for the listed components.

4. Recover the content databases on SQL Server host. ( HERE I CHOOSE MY CONTENT DB and I restore it)

5. Return to the NMM GUI on the web front-end, and click

Continue

in the

dependency dialog box.

Is is a correct procedure or there is something not useful?

I only need a cleat path to follow. If the doc is ok, I'm happy.

1 Rookie

 • 

77 Posts

April 9th, 2012 09:00

Hi,

I'm curious to understand which are the files restored/saved by the WSS writer.

In other words, if I run on a FrontEnd Server nsrsnap_vss_save -G, I can read several xlm files, one for each VSS writer on the system. Therefore I've learned what is saved by IIS configuration writer and IIS metabase, it's clear what is saved by SqlServerWriter on the SQL system, but I cannot figure out what is saved by WSS writer.

Any suggestion please?

Thanks

PierPa

1.7K Posts

April 9th, 2012 23:00

Hi PierPa,

Please note the difference between the XML files (writers information) and the let's say physical files being saved.

To get a clear explanation of the files being saved, or involved for each writer you have to use the DISKSHADOW tool.

This is a tricky one I found out that will help you a lot, specially because the output of these commands can be huge:

You can start diskshadow and logging all outputs to a file by starting a Command Prompt and launching this command:


diskshadow /l c:\diskshadow.out

From the diskshadow prompt, issue the following commands:

   reset
list writers
list writers status
list writers metadata
list writers detailed
exit

The output of all the above commands will be found in c:\diskshadow.out.  The output can use useful to see the list of files which are reported by the VSS writers to be included and excluded by the backup.

The list writers command will show you information in much more detail that vssadmin list writers

list writer status same, will display the information that vssadmin list writers, but in more detail, and also with details about snapshots created, etc.

list writers metadata will show you similar information as if you run nsrsnap_vss_save -G, will provide information about each writer metadata.

And this is the one you really want: list writers detailed. This will display every single file that is saves/used for each writer.

Thank you.

Carlos.

No Events found!

Top