System state synchronizations and express full backups of protected 2012 R2 machines overload DPM server.

I have one DPM 2012 R2 UR2 server which is running on Windows 2008 R2 and 2 protection groups (A, and B). The server uses direct attached storage for all replicas and recovery volumes.
Group A protects the system state and bare metal recovery of 25 Windows 2008R2 servers. It creates recovery points using express full backup each night at 6PM. The servers range from file/print, SQL, domain controllers, terminal servers, web servers, etc. Total
data 3TB.
Group B protects the system state and bare metal recovery of only 3 Windows 2012R2 servers. It creates recovery points using express full backup each night (morning) at 1AM. The servers are 1 domain controller, 1 WSUS server, and 1 server that runs VMware VCenter.
Total data 90GB.
During group A's nightly backup, which takes approximately 2 hours, the DPM server becomes noticeably slower but remains functional. I can open Windows update and check for updates, copy files to and from network locations, launch intenet explorer, etc. It's
stressed since 25 servers are creating backups on it but it remains functional.
During group B's nightly backup the DPM server grinds to a complete halt. Launching internet explorer takes an hour (literally), 10 minutes will pass before I get a menu after clicking the start button. Basically it becomes completely unresponsive until the
backups finish which typically doesn't happen because I'm forced to cancel them upon arriving at work...8 hours later.
I read that DPM uses the built in Windows Backup to protect system state and bare metal recovery. Did something change in Windows Backup from 2008R2 to 2012R2 that would cause this? Why can my DPM server easily backup 25 Windows 2008R2 machines at once in 2
hours yet freezes solid for 8 hours or more when backing up only 3 Windows 2012R2 machines?
I left performance monitor running during the night and it showed an average disk queue length of 49 during group B's backup but everything else was normal. Seems the disks on the DPM server are the bottleneck but again I ask why can my DPM server easily backup
25 Windows 2008R2 machines at once in 2 hours yet freezes solid for 8 hours or more when backing up only 3 Windows 2012R2 machines? Both protection groups use the same disks.
How can I fix this?
Teenage angst has paid of well, now I'm bored and old.

Hi CS,
You can also run a performance monitor.
For some basic perfmon counters to help narrow down the possible bottleneck:
Perf Counters for DPM
Logical Disk/Physical Disk
\%idle
• 100% idle to 50% idle = Healthy
• 49% idle to 20% idle = Warning or Monitor
• 19% idle to 0% idle = Critical or Out of Spec
\%Avg. Disk Sec Read or Write
• .001ms to .015ms  = Healthy
• .015ms to .025 = Warning or Monitor
• .026ms or greater = Critical or Out of Spec
Current Disk Queue Length (for all instances)
80 requests for more than 6 minutes.
• Indicates possibly excessive disk queue length.
Memory
\Pool Non Paged Bytes*
• Less that 60% of pool consumed=Healthy
• 61% - 80% of pool consumed = Warning or Monitor.
• Greater than 80% pool consumed = Critical or Out of Spec.
\Pool Paged Bytes*
• Less that 60% of pool consumed=Healthy
• 61% - 80% of pool consumed = Warning or Monitor.
• Greater than 80% pool consumed = Critical or Out of Spec.
\Available Megabytes
• 50% of free memory available or more =Healthy
• 25% of free memory available = Monitor.
• 10% of free memory available = Warning
• Less than 100MB or 5% of free memory available = Critical or Out of Spec.
Processor
\%Processor Time (all instances)                                                                  
• Less than 60% consumed = Healthy
• 51% - 90% consumed = Monitor or Caution
91% - 100% consumed = Critical
Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread. Regards, Mike J. [MSFT]
This posting is provided "AS IS" with no warranties, and confers no rights.

Similar Messages

  • Understanding replica volume and recovery point volume usage with SQL Express Full Backup

    I am running some trials to test DPM 2012 R2's suitability for protection a set of SQL Server databases and I am trying to understand what happens when I create a recovery point with Express Full Backup.
    The databases use simple recovery model and in the tests I have made so far I have loaded more data into the databases between recovery points since that will be a typical scenario - the databases will grow over time. The database files are set to autogrowth
    by 10%
    I have been looking at the change in USED space in the replica volume and in the recovery point volume after new recovery points and have a hard time understanding it.
    After the first test where data was loaded into the database and an Express Full Backup recovery point was created, I saw an increase in used space in the replica volume of 85 Gb and 29 GB in the recovery point volume. That is somewhat more than I think
    the database grew (I realize that should have monitored that, but did not), but anyway it is not completely far out.
    In the next test I did the same thing except I loaded twice as much data into the database.
    Here is where it gets odd: This causes zero increased usage in the replica volume and 33 GB increased use in the recovery point volume.
    I do not understand why the replica volume use increases with some recovery points and not with others.
    Note that I am only discussing increased usage in the volumes - not actual volume growth. The volumes are still their original size.
    I have been using 3-4 days on the test and the retention period is set to 12 days, so nothing should be expired yet.

    Hi,
    The replica volume usage represents the physical database file(s) size. The database file size on the replica should be equal to the database file size on the protected server.  This is both .mdf and .ldf files.  If when you load data
    into the database and you overwrite current tables versus adding new ones, or if there is white space in the database files and the load simply uses that white space, then there will not be any increase in the file size, so there will not be any increase
    in the replica used space.
    The recovery point volume will only contain delta changes applied to the database files.  As the changed blocks overwrite the files on the replica during express full backup, VSS (volsnap.sys) driver copies the old blocks about to be overwritten
    to the recovery point volume before allowing the change to be applied to the file on the replica. 
    Hope this helps explain what you are seeing.
    Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread. Regards, Mike J. [MSFT]
    This posting is provided "AS IS" with no warranties, and confers no rights.

  • DPM Express Full Backups Causing SQL timeouts

    Hi,
    We run a SQL 2008 R2 failover cluster on Server 2008 R2 SP1 connected to an Equallogic SAN. The SQL failover cluster is fully servicepacked and installed the latest cumulative updates. 
    We have all our databases in full recovery mode and protected by DPM 2012 SP1 with latest cumulative updates.
    The protection groups for SQL are configured to synchronise every 60mins and 1 express full backup every night. Every time the SQL express backups run the SQL server experiences timouts and end users report applications timing out. In the SQL server Application
    logs I see Event ID 833 numerous times, these only occur as the express full backups take place. Every time the express full backups run we see the same timeouts and event logged that I/O has taken longer than 15 seconds to complete:
    Log Name:      Application
    Source:        MSSQLSERVER
    Date:          01/04/2014 04:10:10
    Event ID:      833
    Task Category: Server
    Level:         Information
    Keywords:      Classic
    User:          N/A
    Computer:      WDCSQL02.local
    Description:
    SQL Server has encountered 1 occurrence(s) of I/O requests taking longer than 15 seconds to complete on file [D:\MSSQL10_50.MSSQLSERVER\MSSQL\DATA\MSDBLog.ldf] in database [msdb] (4).  The OS file handle is 0x00000000000008F8.  The offset
    of the latest long I/O is: 0x000000042eae00
    Event Xml:
    <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
      <System>
        <Provider Name="MSSQLSERVER" />
        <EventID Qualifiers="16384">833</EventID>
        <Level>4</Level>
        <Task>2</Task>
        <Keywords>0x80000000000000</Keywords>
        <TimeCreated SystemTime="2014-04-01T03:10:10.000000000Z" />
        <EventRecordID>14165629</EventRecordID>
        <Channel>Application</Channel>
        <Computer>WDCSQL02.local</Computer>
        <Security />
      </System>
      <EventData>
        <Data>1</Data>
        <Data>15</Data>
        <Data>D:\MSSQL10_50.MSSQLSERVER\MSSQL\DATA\MSDBLog.ldf</Data>
        <Data>msdb</Data>
        <Data>4</Data>
        <Data>00000000000008F8</Data>
        <Data>0x000000042eae00</Data>
        <Binary>410300000A0000000600000055004B00530051004C00000000000000</Binary>
      </EventData>
    </Event>
    I was under the impression that DPM SQL backups would remain online during the backup.
    Is this normal? 
    Any way to fix this issue?
    Thanks,
    Microsoft Partner

    Hi
    A copy backup will not truncate the logs but a full backup will and it will also stamp the database once it does a full backup.

  • Maximum Express full backups per day per Datasource?

    I have 3 DB server having 24 databases. Some of these are SIMPLE recovery model and some are FULL. Express full backups are set one in a day at 8:00 PM and Synchronization is set to every 4 hours.
    Now customer want to keep backup of every 2 hours. So I have some queries on this.
    1) Its not possible to set EXPRESS full backup after 2 hours. So what should I do here, Should set synchronization frequency to 2 hours? 
    2) If I set Synchronization Frequency to 2 hours, it will backup only FULL Recovery model databases only? SIMPLE one will be backed up 8 PM only?
    3) Incremental Backups recovery is only allow to restore Instance? cant be copy to Network folder? As when I tried to restore 4 hours synchronization backup it shows cant be copied, only recovery point associated with express backup will be copied.
    4)Based on my understanding, I should change all recovery models to FULL and should set Synchronization to 2 Hours to achieve customer demand. It will allow me to restore 2 hours recovery point to directly Instance or another instance (But not to Network
    folder and tape in case of Incremental recovery) 
    Am I right? Thoughts!
    Thanks in Advance
    VJay

    See.
    My 4 hours Synchronization frequency/Incremental backups are creating recovery points successfully and I can see these from recovery tab. But when I am trying to restore Incremental recovery point, it doesn't allowing me to restore. See in below image, I
    am trying to restore Incremental recovery point which was created yesterday at 4:00 PM but its asking me to restore yesterday's 12:20 PM. This incremental backups tooks place after 12:20, doesn't mean anything in terms of recovery!
    As my final decision, should I choose EXPRESS Full backup for evry 2 hours and Synchronization just before Express backup for these 24 databases since these are now in FULL model.

  • DPM is Only Allowing Express Full Backups For a Database Set to Full Recovery Model

    I have just transitioned my SQL backups from a server running SCDPM 2012 SP1 to a different server running 2012 R2.  All backups are working as expected except for one.  The database in question is supposed to be backuped up iwht a daily express
    full and hourly incremental schedule.  Although the database is set to full recovery model, the new DPM server says that recovery points will be created for that database based on the express full backup schedule.  I checked the logs on the old DPM
    server and the transaction log backups were working just fine up until I stopped protection the data source.  The SQL server is 2008 R2 SP2.  Other databases on the same server that are set to full recovery model are working just fine.  If we
    switch the recovery model of a database that isn't protected by DPM and then start the wizard to add it to the protection group it properly sees the difference when we flip the recovery model back and forth.  We also tried switching the recovery model
    on the failing database from full to simple and then back again, but to no avail.  Both the SQL server and the DPM server have been rebooted.  We have successfully set up transaction log backups in a SQL maintenance plan as a test, so we know the
    database is really using the full recovery model.
    Is there anything that someone knows about that can trigger a false positive for recovery model to backup type mismatches?

    I was having this same problem and appear to have found a solution.  I wanted hourly recovery points for all my SQL databases.  I was getting hourly for some but not for others.  The others were only getting a recovery point for the Full Express
    backup.  I noted that some of the databases were in simple recovery mode so I changed them to full recovery mode but that did not solve my problem.  I was still not getting the hourly recovery points.
    I found an article that seemed to indicate that SCDPM did not recognize any change in the recovery model once protection had started.  My database was in simple recovery mode when I added it (auto) to protection so even though I changed it to full recovery
    mode SCDPM continued to treat it as simple. 
    I tested this by 1) verify my db is set to full recovery, 2) back it up and restore it with a new name, 3) allow SCDPM to automatically add it to protection over night, 4) verify the next day I am getting hourly recovery points on the copy of the db. 
    It worked.  The original db was still only getting express full recovery points and the copy was getting hourly.  I suppose that if I don't want to restore a production db with an alternate name I will have to remove the db from protection, verify
    that it is set to full, and then add it back to protection.   I have not tested this yet.
    This is the article I read: 
    Article I read

  • Express Full Backups

    Good Afternoon,
         I am thinking about setting my express full backup to run during the day during the day to decrease supporting this at night. My question is will this impact our databases to the point of customer impact. We currently do 4 sync jobs a
    day, then a full backup on Friday night. Just trying to make sure there is no impact.
    Thanks

    It will also depend on how you are setup, if you have a DAG you can run a full backup against the passive copy. I would suggest doing an EF once a day at night and sync to what your recovery goal needs are.
    Regards, A.Nadar, This posting is provided &quot;AS IS&quot; with no warranties, and confers no rights.

  • Currently have 30K pics on PC Win. 7 laptop using  PSE 8, corrected issues and did full backup on portable hard drive plus saved (Extensive) Keyword tags to file.  Installed PSE 13 on PC Win 8.1 laptop.  Will attach same portable Hard drive to new laptop.

    Currently have 30K pics on PC Win. 7 laptop using  PSE 8, corrected issues and did full backup on portable hard drive plus saved (Extensive) Keyword tags to file.  Installed PSE 13 on PC Win 8.1 laptop.  Will attach same portable Hard drive to new laptop. How do I import prev Keyword Tags from file and Restore catalog from port. HD to new PC and in what order?  I have not opened PSE 13 on new PC yet. Would like to do it right the first time!

    No importing needed.
    See these instructions: Use Backup, Restore to move catalog | Organizer | Elements 6 or later

  • Compare system state (number and types of objects in memory)

    We have 2 exactly identical servers running our application talking to the same database.
    When some obejct from server 1 is deleted, a message is passed to server 2 to delete the object from its cache too.
    Forget the database, I want to design a testing system which compares the system-state (number and type of live-objects) in memory for server 1 and server 2 after the operation and the message, and reports any discrepancy.
    Any ideas ?

    Are you trying to debug a distributed cache implementation?
    If you want to inspect the state of your application on multiple servers creating a JMX bean might be a solution.
    http://java.sun.com/developer/technicalArticles/J2SE/jmx.html

  • Keeping ARCH and Deleting FULL backups on RMAN

    Dear all,
    RMAN deletes all my full backup and archivelogs in 2 days with;
    DELETE NOPROMPT OBSOLETE RECOVERY WINDOW OF 2 DAYS;
    But I want to keep archivelogs at least 9 DAYS and delete just the FULL backups after 2 DAYS. How can i set this?
    Any recommendations?
    Regards

    You reported the Sybrand's post as an "abuse to consider" to the mod. Not to take a part or the other, but Sybrand made several points you should yourself to consider up front :
    1. you claimed your problem as an emergency case, it is considered as rude in a public forum based on volunteers.
    2. also you did not provided any information which can make difference, as the minimum of Oracle version (and OS...), for one who want to be considered as an emergency, that's rather strange, no one here is sitting next to you
    3. Sybrand gave you a command to run that you should see the suggestion resulted
    4. He also made a point about the documentation, whether it can appears rude to you, it's always, ALWAYS, good to show exactly to your peers what you have done so far to solve your problem, based on what docs...
    5. We could also point you a link such as How to ask question
    Please, keep it in mind, and come down.
    Nicolas (moderator).
    PS: without saying that 2 mods spent their time to proceed with your faulty reported abuse case
    Message was edited by: Nicolas.Gasparotto

  • Time Capsule and iTunes full backup problem

    Hi,
    Every time I add something to my iTunes library the time capsule does a full backup of the entire library rather than just doing an incremental backup.
    This has only been happening for the last month or so but it is very annoying and has resulted in my 2TB time capsule starting to delete older backups!
    Any ideas why this might be happening and any suggestions on how to solve it?
    I could delete all the previous backups and start again but would prefer not to have to do that.
    Thanks,
    Anthony

    have a look at this user tip.

  • After a complete restore of Mac from a Time Machine volume, I don't want a full backup the next time Time Machine backs up.

    After performing a complete restore of my Mac from a Time Machine volume, Time Machine wants to perform a full backup the next time it backs up.This is normal: http://support.apple.com/kb/TS1338
    But I don't have enough free space on my external hard drive to do this because my external hard drive already contains one complete backup. My hard disc on my MacBook is 320 and my external drive is 500 GB, so it doesn't add up. I don't want to delete the original backup in case something goes wrong during the new full backup. What can I do?

    You don't have too many options.  The best approach would be to buy a larger external drive.  The guideline for Time Machine is that you should have 2-3 times as much capacity on the external drive as data you're backing up.  See item 1 of Time Machine - Frequently Asked Questions.  In any case, you can't prevent that first full backup unless you're willing to turn off Time Machine permanently.

  • My system was re-set during repairs and I have lost my Firefox bookmarks. Are these stored anywhere on the system? I did a full backup of documents and programme files before repair, but can't find any bookmark file.

    I did not do backups of the bookmarks regularly, since I thought these would be saved somewhere. However, I can't find any bookmark files on my backed-up system.

    Thanks - now I can see an applications folder, and have found a .default file, hope this is the one. Should I now follow these instructiosn to restore my profile, or just go to organise bookmarks and import the default file?
    Copying files between profile folders
    Note: Firefox must be closed down while copying files between profile folders.
    1. Open your profile folder:
    * (Firefox 3.0 and 3.5): See How to find your profile
    1. (Firefox 3.6): At the top of the Firefox windowOn the menu bar, click on the Help menu and select Troubleshooting Information.... The Troubleshooting Information tab will open.
    2. Under Application Basics, click on Open Containing FolderShow in Finder.
    2. From the menu at the top of the Firefox windowbar, select FileFirefox and then select the ExitQuit FirefoxQuit menu item.
    3. Right-clickHold down the Ctrl key while you click on a file you want to copy and choose Copy.
    4. Open the new Firefox profile folder.
    5. Right-clickHold down the Ctrl key while you click inside the profile folder and choose Paste Item.
    Repeat steps 2-5 for each file you wish to transfer over to your new profile folder.
    While copying your personal data from an old Firefox profile to a new Firefox profile, it is possible that you could copy over the file that is causing the problem you are trying to get rid of! The more files you copy over, the greater the chance of this happening. It is therefore recommended that:

  • System stats - files and memory

    Please accept my apology if this is the wrong forum!
    Is there a way - system calls eg. - to tell how much memory and how many files a process or thread uses? I know I can use /proc to find info about file handles, but I expect there must be a place where the system keeps this information?
    I want to be able to tell from within the program itself - I'm using third party libraries, and apparently they allocate inordinate amounts of resources without freeing them; and they are not about to admit the fault is in their code, so I have to see which resources are in use before and after calling their code.

    Please accept my apology if this is the wrong forum!Yeah, this is Solaris OS question :-)
    Is there a way - system calls eg. - to tell how much
    memory and how many files a process or thread uses? I
    know I can use /proc to find info about file handles,Ugh, what's wrong with it?
    but I expect there must be a place where the system
    keeps this information? It's kept by kernel. The only interface you can see it is prosfs
    or proc tools (man -s1 proc), but these tools query /proc
    anyway,

  • If I filled the system hard drive, and I'll delete anything on it, it's off and on the external drive, or can I safely clean the system hard drive and all that was done in the Time Machine in the same place, on an external hard drive and will remain ?

    Please help

    Time Machine does delete "old" backups, eventually.  "Eventually" is controlled by whether or not you have used all of the space on the TM drive.  If you have "more free space" on the TM drive, then nothing should be deleted yet.
    If youd elete files from a system drive, and continue on the same TM set, TM does NOT scour all old records to purge missing files ... else the "time" part of "Time Machine" makes no sense.

  • Problem with full backup with copy only in maintenance plan sql server 2008

    Hello
    I am starter DBA for sql server in general
    I have problem related to backup with maintenance plan in sql server 2008 and 2008 r2
    I created maintenance plan to take full back up  with copy only option checked, but when I check the back up in backupset table is_copy_only column it shows 0 for databases that I run job for.
    I did the same steps on sql server 2012 and I checked backupset table is_copy_only column it shows 1 for databases that I run job for .so it works as it should be
    Note: all the servers I mention here are production servers 

    Copy-only Backup  by using GUI ( SSMS) in SQL Server 2008 but it was not available in  Maintenance Plan "Back
    Up Database Task". Now in SQL Server 2012 It is included in "Back Up Database
    Task". 
    http://sqlage.blogspot.in/2013/06/dba-maintenance-plan-back-up-database.html

Maybe you are looking for