Partager via

My DPM article on 'You Had Me At EHLO' generated a number of comments...

  • Article

I recently posted an article to the Exchange Team site here - 'Protecting Exchange data with Microsoft System Center Data Protection Manager (DPM)' and received a number of comments. I thought I would discuss some of the issues here on my blog rather than with such a wide audience on 'You Had Me At EHLO'.

It seems the reasons given to the implement CCR are also be used for DPM. I do know of a few shops leveraging CCR as a solution in both HA and backup space.
March 20, 2008 12:09 PM 

I have been involved in a number of ongoing discussions where there does appear to be a lot of overlap between how and why you would choose DPM or any other backup solution over or alongside CCR, SCR, mailbox retention, deleted item retention etc etc etc...  My advice would be to get all interested parties together and discuss exactly what you need and what you can realistically afford to deploy; come up with a first draft design and then decide where your design does or doesn't meet your requirements; then come up with a better design where appropriate.  For example, one that I've come across is using DPM Primary and Secondary servers for site resilience alongside SCR.  Why use DPM primary and secondary and SCR?  A lot of overlap and you need a lot of bandwidth to replicate the same data twice via two different mechanisms.

To answer the original question CCR would be used as mitigation against server or server component failure, storage failure, physical database corruption and in some cases site resilience.  DPM would generally be used to provide protection against different things - logical database corruption, long term data retention and possibly administrator error. There is certainly overlap there as to what CCR and\or DPM provides protection against.  However one thing that is worth considering is the minor risk that logical database corruption occurs and is replicated to both copies of your CCR database.  Having the option of a DPM recovery point would mitigate against this.  However the other options in the event of logical corruption are to create a new database and move mailboxes or to use isinteg to repair the corruption....?  I believe it is all about evaluating risk against impact against SLA and finally cost and then designing appropriately. Have a look at this post I put together a while ago - might be interesting? (Do we actually need to backup our Exchange data anymore?)

The backup functionality works great, now let’s look at recovery. I believe people will find that without third party apps and a huge amount of time and process recovering mailbox’s is a painful and unrealistic process.Please check official MS documentation. Once again the MS world and the real world are two different things, WHY? Don’t get me wrong I’ve been with DPM for awhile now, cut my teeth on Beta. What DPM “is”, “What MS thinks DPM is” and what DPM "could be" are three very different things unfortunately.I hope this gets looked at and corrected soon.
March 31, 2008 11:01 AM

I am honestly quite surprised at this comment.  I don't understand what is so painful and unrealistic about the recovery process.  If you are talking about individual mailbox recovery then the process is to recover an entire database to recovery storage group or recovery server and extract the appropriate data. OK so it's not a one-click process but if you are restoring mailboxes on a regular basis then it is my guess that there is some process or operation that is not working.  I don't think there should necessarily be a technical resolution to this. In my opinion in a properly designed infrastructure deployment you should only ever restore from backup for the purposes of a firedrill to test that your procedures and scripts work and to ensure that can continue to meet your SLA's.  I honestly don't believe that that is unrealistic, even taking into account unexpected outages that will always occur in even the most robust deployments.

I think you are suggesting that DPM should use some sort of bricks-level backup to provide single click mailbox restore.  This has a number of implications; the first being the expense time, storage and performance of using MAPI to log into every mailbox to backup each mailbox item by item, the second being the point at which mailbox restores become part of your daily routine as opposed to an option of last resort.  I would love to hear your comments because I might have missed the point completely!

Great article on how DPM and Exchange work, and can work together. Unfortunately, Milo's comment hits the real problem right on the head. DPM's strengths and abilities are greatly diminished with the Exchange recovery steps. Please see discussions in Microsoft Communities on Data Protection Manager. We're still waiting on real answers from Microsoft.
March 31, 2008 11:37 AM

Thanks for the comment.  Will take a look at the Communities.  Please see comments above.

Doug Gowans from the Microsoft Premier Support Team in Dubai has posted a detailed article summarizing how DPM2k7 functions in relation to Exchange 2007. What I find of value in this Blog Post is the summarization of how DPM2k7 incorporates Volume Shadow Copy (the VSS Writer) function into its inherent Backup function.
March 21, 2008 10:13 AM 

I'm not in the Premier Support Team in Dubai but thanks for the plug!  (I work in MCS in the UK now.)

Thank you very much. the diagrams really made it easy for me to understand. by the way, DPM at present does not support windows server 2008, right?
March 21, 2008 10:22 AM 

Thanks for the comment. So at the moment you are correct "This release of DPM 2007 does not support a Windows Server 2008 installation in a production environment".  Take a look here for more information.  Having said that it is my understanding that support for Windows 2008 is very close and will be announced soon.

Great Post! Some questions: How does the feature "Enable SAN based recovery using hardware snapshots" interact during DPM restore? What are the best practices for Exchange DB and Log volumes in this sense? Are in this case all the snapshots (incremental and full) not discarded? Are only SANs supporting the Virtual Disk Service compatible with this model?
March 20, 2008 10:56 AM

Thanks for the comment.  I have to admit to not having any direct experience of using this option.  It is my understanding however that this option is used when you have your DPM server and Exchange server supported by the same SAN.  My guess that this hands control of the storage from the default Windows System VSS Provider over to that provided by the SAN vendor.  Database and Transaction Log recommendations would remain the same but the rest of the question I cannot answer unfortunately. I will find out and blog again...

How does this stack up against a solution that utilizes host-based replication for protecting Exchange like Double-Take?
March 24, 2008 4:01 PM

I have limited experience of Double-Take but it is my understanding that it is more similar to continuous replication technology using data replication as opposed to block level volume snaps with DPM. I believe you would pitch Double-Take against CCR or SCR as opposed to DPM in which case if I was deploying Exchange 2007 I would want the integration and level of support from Microsoft that CCR and SCR would provide.  I think it comes down again to what you want from your solution and then design appropriately using backup, continuous replication etc etc.. (Have a read of my blog if you are interested in reading more about why you would or wouldn't use DPM and\or continuous replication.)

This is great info, but can you give us an article about restoring from DPM and what can be done? Backups are great, but what really matters is restore.
March 25, 2008 12:01 PM

Thanks for the comment - yes I hope there will be some follow up articles - I have begun jotting some ideas down already.

Comments