DOCUMENTATION: Description of Symantec NetBackup Disk Staging Relocation Behavior

Article:TECH44719  |  Created: 2005-01-08  |  Updated: 2010-10-02  |  Article URL http://www.symantec.com/docs/TECH44719
Article Type
Technical Solution


Environment

Subject

Issue



DOCUMENTATION: Description of Symantec NetBackup Disk Staging Relocation Behavior


Solution



I. Introduction
One of the features of NetBackup is the ability to utilize more than one drive for a single Disk Staging Storage Unit (DSSU) relocation job. This document is an explanation of how DSSU's works, and how this behavior can be tuned.
 


    The default behavior in NetBackup is for a DSSU relocation job to first divide its list of images to be duplicated into separate groups for each retention level represented in the list.  Each of these retention groups is separated into individual batches that do not exceed 25 gigabytes of data per batch. Next, a bpduplicate command is executed to handle each batch of images.  If enough drive resources are available, all of these duplication jobs will immediately activate.  Otherwise, there will be a mix of active and queued duplication jobs based on drive availability.

    All of this batching behavior is the responsibility of the bpbrmds process that runs on the media server that originally wrote to the DSSU.  For debugging purposes, the bpbrmds process will write logging information to files in the bpbrmds log directory on the media server, if this directory is present.

    For the examples given below, consider a hypothetical NetBackup environment with a single master/media server that includes a single tape library with ten tape drives.  This environment is configured with the following pair of storage units:

        Table 1: Example Configuration - Two Storage Units

        Storage Unit Name Storage Unit Type Max Concurrent Write Drives Final Destination Storage Unit
        dlt-robot0 Media Manager 10 (NA)
        dssu Disk Staging (NA) dlt-robot0


      Example 1:
      One hundred images have been written to the DSSU disk staging storage unit.  Each image is 1 gigabyte in size. Fifty of these images were configured with retention level 1, and fifty of these images were configured with retention level 2.  

      If a relocation job is launched for the DSSU storage unit, the result would be an immediate launching of four duplication jobs in the activity monitor:
      • First and Second jobs are launched to duplicate all 50 images for retention level 1  (25 images per job)
      • Third and Fourth jobs are launched to duplicate all 50 images for retention level 2 (25 images per job)
      •  

      Example 2:
      One hundred images have been written to the DSSU disk staging storage unit.  Each image is 1 gigabyte in size.  Twenty of the images were configured with retention level 1, and eighty of these images were configured with retention level 2.

      If a relocation job is launched for the DSSU storage unit, the result would be an immediate launching of five duplication jobs in the activity monitor:
      • First job is launched to duplicate all 20 images for retention level 1
      • Second, third, and fourth jobs are launched to duplicate the first 75 images for retention level 2 (25 images per job)
      • Fifth and final job is launched to duplicate the remaining 5 images for retention level 2

      Example 3:
      Three hundred images have been written to the DSSU disk staging storage unit.  Each image is 1 gigabyte in size.  Half of these images were configured with retention level 1, and half of these images were configured with retention level 2.

      If a relocation job is launched for the DSSU storage unit, the result would be an immediate launching of twelve duplication jobs in the activity monitor:
      • First, second, third, fourth, fifth, and sixth jobs are launched to duplicate all 150 images for retention level 1 (25 images per job)
      • Seventh, eighth, ninth, and tenth job is launched to duplicate the first 100 images for retention level 2 (25 images per job)
      • The eleventh and twelfth jobs would also be launched (to duplicate the last 50 images for retention level 2), but these two jobs would remain queued until drive resources are available

      •  
    III. How this Behavior can be Modified
    Some NetBackup administrators may not be satisfied with the above behavior.  Administrators may want to minimize the number of new tapes that are used to hold the duplicated images, which is best done by increasing the maximum amount of data that is duplicated by a single duplication job.   Additionally, some administrators may want to guarantee one or more idle drives during DSSU relocation to handle any unexpected backup or restore need.  This can be done by decreasing the number of duplication jobs that will be simultaneously activated during the DSSU relocation operation. Instructions on how to achieve these changes are described below.

      III.A. How to Change the Maximum Amount of Data in a DSSU Duplication Job
      By default, a DSSU relocation job will limit the size of duplication jobs to no more than 25 gigabytes of data.

      The bpbrmds process will look for a text file named STAGING_JOB_KB_LIMIT in the NetBackup directory on the media server that originally wrote to the DSSU.  This text file should contain a single line with a single value: the desired data size limit in kilobytes.  This value can range from 1024 (1 megabyte) to 2147483647(2 terabytes).

      It is not recommended to set the value in STAGING_JOB_KB_LIMIT to greater than 99614720KB if running multiple copy staging jobs.  See TechNote ID 338561 in the related documents for details on why this limit exists


      For example, to configure NetBackup to allow DSSU duplication jobs to occur in batches of up to 50 gigabytes per batch, the following command should be run:

          UNIX media server:
          echo 52428800 > /usr/openv/netbackup/STAGING_JOB_KB_LIMIT

          Windows media server:
          echo 52428800 > %systemdrive%\Program Files\VERITAS\NetBackup\STAGING_JOB_KB_LIMIT

      III.B. How to Decrease The Number of Simultaneous Duplication Jobs That Will be Activated
      As seen in Example 3 above, it is possible for a DSSU relocation job to create enough duplication jobs to occupy all the tape drives in a tape library.  It is possible to configure NetBackup to activate a smaller number of duplication jobs for a DSSU relocation job.

      To do this, the administrator should define an additional media manager storage unit that is configured with a Max Concurrent Write setting that is equal to the maximum number of drives that should be used at one time for DSSU relocation. The customer's DSSU configuration should then be modified to specify this new storage unit as the FSU.  (see section III.C. below for an example)

      III.C. Example 3 Revisited, Applying these Modifications
      To illustrate these two modifications with an example, once again consider our hypothetical environment of a single tape library with ten tape drives.

      The STAGING_JOB_KB_LIMIT text file is present on this example master/media server, and this file contains "52428800" to represent a 50 gigabyte duplication job size limit.

      Instead of the configuration shown in Table 1, this example includes the storage unit configuration shown in Table 2:

          Table 2: Example Configuration - Three Storage Units

          Storage Unit Name Storage Unit Type Max Concurrent Write Drives Final Destination Storage Unit
          dlt-robot0 Media Manager 10 (NA)
          dlt-robot0-5drives Media Manager 5 (NA)
          dssu Disk Staging (NA) dlt-robot0-5drives


      Repeating the scenario of Example 3 (300 images, each 1 gigabyte, split evenly between two retention levels) will produce slightly different results in this configuration:
      • First, second, third duplication jobs are launched to duplicate the 150 images for retention level 1 (50 image per job)
      • Fourth, fifth, and sixth duplication jobs are launched to duplicate the 150 images for retention level 2 (50 images per job)
      • Assuming drive availability, only the first five jobs will activate.  The sixth job will remain queued until one of the first five jobs is complete.

      Drives
    IV. Additional Notes
    In the case where a DSSU relocation job encounters a single backup image that is larger than the duplication job size limit, a duplication job will be created to duplicate only that single image.


II. Default NetBackup Behavior When Relocating Images from DSSU to FSU
 




Legacy ID



280067


Article URL http://www.symantec.com/docs/TECH44719


Terms of use for this information are found in Legal Notices