none
'Tape Backup' jobs - Can you determine what 'disk backup' point it is using (going to use) within teh 'Jobs' menu or via an SQL script RRS feed

  • Question

  • Hi

    I would liek to know if it is possible to see the date and time of the disk backup point that the 'Tape Backup' job is\will be using while it is running\queuing in the Monitoring\Jobs section when 'All jobs in progress' id teh filter option?

    The reason is so that when we get a Tape backup failure which puts us hours\days behind (which DPM will not be able to catch up with) I can make a confident decision on what I can afford to live without.

    Thanks

    Wednesday, February 20, 2013 4:32 PM

All replies

  • Hi,

    Long term tape backups always mount the latest disk based recovery point (snapshot) and copy that to tape.   All long term tape backups are full backups, so if the data source is exchange or SQL, that would be the last express full backup time.  For all other data sources, that would be latest successful sync prior to the recovery point (shadow copy). 

    So, there is no easy way to determine as seen in the example below.

    Suppose you do express full backups at 8:00PM but tape backups run at 10:00PM - then all tape recovery points will show 8:00PM.

    Suppose all other data sources use 30 minute synchronizations and get RP's made at 8:00am, 12:00PM, and 6:00PM - then tape recovery point will show 6:00PM UNLESS a Synchronization failed prior to the 6:00PM recovery point.  lets say the last successful Synchronization was actually at 4:00pm for a protected Client , and the recovery point was made at 6:00PM, then the tape backup would show 4:00PM for that client, while other clients may show 6:00PM. 

    In a nutshell, tape backup recovery point time will mimic latest disk based recovery point time, which is actually latest synchronization time prior to the snapshot.   Under recovery tab, the time of recovery point is always that of the last time the data was brought over to the DPM server prior to recovery point time. 


    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.

    Wednesday, February 20, 2013 10:08 PM
    Moderator
  • Hi Mike

    Thanks for the explanation. The reason I want to know if this is possible is because when we have a failure (i.e. DPM fails to sync with the tape library) which causes a number of failed tape jobs and both of our tape drives then become inactive for a while. In an ideal world we would just restart the tape jobs and then hope that they would all run through and complete however our current status means that our 2 tape drives are only idle for 2-3 hours per day (usually from 9pm to 11:59pm) although that is when everything is running smoothly. As DPM does not support automatic tape cleaning the tape cleaning issues normally put us back a few hours each week (as both drives usually need cleaning once a week at randon times - normally in the early hours of the morning) and need to be manually actioned which is not always possible to action quickly.

    With the above in mind after a tape library sync fail, when the failed backup jobs are restarted (bearing in mind that soem of these jobs may now be 2-3 days old) all of the jobs will now show a start date of when they were restarted and not the original start date. This makes it impossible to tell which 'disk backup date' the job is from. You can only find 'disk backup date' information out once the job is running and the tape is in one of the drives (right click tape and 'View Tape Contents') which is too late as valuable time and resources have then been consumed.

    As you can see, this is why I need to know the 'disk backup date' and other than getting a 3rd tape drive to help with the tape backup load (not an option) I cannot think of anything more I can do to help me to analyse the risk and make an educated call on what I can safely write off.

    Thursday, February 21, 2013 11:43 AM
  • Hi,

    When you restart a failed backup job from the alerts tab, that tape backup will use the original recovery point date/time, not the latest. Are you saying some tape backup jobs are more valuable to you than others, and you would like to cancel less important ones based on recover point time ?


    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.

    Thursday, February 21, 2013 2:01 PM
    Moderator
  • That is correct Mike.

    The problem we have is that, because once we have a failure that puts us behind (and with the cleaning issue) DPM does not have the time to catch up (as before it even gets close another failure will occur due to 1 or both drives needing cleaning). This then causes the 2 tape drives to be constantly running for 24hrs for days\weeks. We have been in this situation for a while now and are trying to find a way to manage it as best we can.

    Our 'disk backups' stay available for 5 days so if DPM continues to get behind with the Tape jobs at some point it will not be able to take a tape backup because there will be no 'disk backup' available as it has expired. This is why we need to make a decision on the facts of what we can see from DPM rather than leaving it as long as we can and then 'binning off' all tape jobs and starting again which leaves us more exposed than 'cherry picking' for want of a better expression. We have had to 'bin off' all jobs already and I would like to avoid this action a second time if possible.

    The long term solution for us involves moving from DPM as we realise that it cannot handle what we need with our current hardware however it is likely that we will utilise DPM somewhere else in our business that does not require as much data to be backed up. This is why I am exploring other avenues as the business will not spend money on DPM when it is planned to be decomissioned in the next 3-6 months.

    There must be some way of seeing this in SQL if not via the DPM Console or am I chasing a lost cause?

    Thursday, February 21, 2013 3:08 PM