Backup source database

Source database with Raw DEVICE type storage

Several users use raw device-based tablespaces for source DB2 databases. To leverage these environments with Delphix, Delphix has built a workflow using Db2s native tools that allow Delphix to discover and convert a raw device-based tablespace into an automatic storage-based tablespace during ingestion. Once the data is ingested into staging, customers will be able to provision VDBs of the automatic storage-based database. 

In order to complete the linking process, the Standby dSource must have access to a full backup of the source Db2 databases on disk. This should be a compressed online DB2 backup and must be accessible to the dSource instance owner on disk. Delphix is currently not setup to accept DB2 backups taken using third-party sources such as Netbackup or TSM. All three features of data ingestion, namely HADR, Non-HADR, and Customer Supplied Archive Logs backups must also include logs. Starting with Db2 plugin version 3.1.1, we are also supporting full online backups, which are taken using the "exclude logs" syntax. This support is only applicable to the dSources created using the Customer Supplied Archive Logs (Backup and Logs) ingestion mechanism on a single partition.

Example backup command: db2 backup database <DB_NAME> online compress include logs

Best practices for taking a backup

The following best practices can help improve backup and restore performance:

1. Compression should be enabled

2. Following parameters should be optimally configured:

   1. Utility Heap Size (UTIL_HEAP_SZ)

   2. No. of CPUs

   3. No. of Table Spaces

   4. Extent Size

   5. Page Size

3. Parallelism & Buffer configuration may be used to improve backup performance. Parameters that should be configured are :

   1. Parallelism

   2. Buffer Size

   3. No. of Buffers

More information about backup best practices is available in IBM Knowledge Center

Procedure

1. Login to the Delphix Management Application using Delphix Engine credentials or as the owner of the database from which you want to provision the dSource.

2. On the Databases tab of Environment Management screen, add a source config against discovered staging instance.

3. Then, click Manage.

4. Select Datasets.

5. Click the Plus (+) icon and select Add dSource, you’ll get a list of available source configs using which you can go for dsource creation.

6. In the Add dSource wizard, select the required source configuration.

7. If you are working with an HADR setup, please leave the HADR checkbox checked unless you're working with Staging push dSource.

8. The database name is mandatory and must be unique for a given instance. This is the name that the database was on the instance it was restored from.

9. Enter the complete Backup Path where the database backup file resides. If the field is kept blank, the instance home directory will be used as the default value. If there are multiple backup files for a database on the backup path, the latest one will be used. If the backup files are split across multiple filesystems, you can provide them as additional backup paths with the first path being populated in the Backup path field.

10. Enter the Log Archive Method1 you wish to use for the database. If no value is entered, the default value used is DISK:/mountpoint/dbname/arch

11. Optionally, users can set the database configuration parameters during the linking operation in the Config Settings section.

12. If the dSource is to use HADR please enter the following fields. If it will not use HADR skip ahead to step 13. For more information about HADR please view Linking a dSource from a Db2 Database: An Overview

    1. Enter a fully qualified HADR Primary Hostname. This is a required field for HADR and must match the value set for HADR_LOCAL_HOST on the master.

    2. Enter the port or /etc/services name for the HADR Primary SVC. This is a required field for HADR and uses the value set for HADR_LOCAL_SVC on the master.

    3. Enter the port or /etc/services name for the HADR Standby SVC. This is a required field for HADR and uses the value set for HADR_REMOTE_SVC on the master.

13. Click Next.

14. Select a dSource Name and Database Group for the dSource.

15. Click Next. You will get Data Management section where you need to specify staging environment and user which will be used for dsource creation.

16. Set the Staging Environment to be the same as the dSource host.

17. Select the Staging Environment User to be the same as the instance owner of the dSource instance. Info: Changing the Environment User

    If you need to change or add an environment user for the dSource instance, see Managing Db2 Users and Instance Owners

18. Then, click Next and you’ll get the Policies section. Set the desired Snapsync Policy for the dSource.

19. Click Next.

20. Specify any desired pre- and post-scripts. For details on pre- and post-scripts.

21. Click Next.

22. Review the dSource Configuration and Data Management information in the summary section.

23. Click Submit.

The Delphix Engine will initiate two jobs to create the dSource: DB_Link and DB_Sync. You can monitor these jobs by clicking Active Jobs in the top menu bar, or by selecting System > Event Viewer. When the jobs have completed successfully, the database icon will change to a dSource icon on the Environments > Host > Databases screen, and the dSource will also appear in the list of Datasets under its assigned group.

The dSource configuration screen

After you have created a dSource, the dSource Configuration tab allows you to view information about it and make modifications to its policies and permissions. In the Datasets panel, select the dSource you wish to examine. You can now choose the configuration tab to see information such as the Source files, Data Management configuration, and Hook Operations.