Reducing Load on the IBM Db2 Server

To reduce the load that backup operations put on the IBM Db2 server, you can throttle Veeam Plug-In performance and move digest calculation to the backup repository. Configure performance throttling first, and move digest calculation to the backup repository only if throttling does not reduce the CPU load enough.

Configuring Performance Throttling

You can limit computer processing (CPU) resources assigned for Veeam Plug-In backup jobs. This setting allows you to reduce the impact of backup operations on the target computer performance. Performance throttling prevents Veeam Plug-Ins from utilizing all computer resources to ensure that enough resources are provided for other operations. This may be useful in case you detected that your source database responses are slower during backup operations.

If you throttle performance, additional system components help with the throttling process:

• For Linux or Unix, Veeam Backup & Replication starts Veeam Plug-In backup operation processes with low priority and nice value of 19. The nice values are in ascending order of priority ranging from -20 (the highest priority) to 19 (the lowest priority).
• For Microsoft Windows, Veeam Plug-In uses the OS process priority system in the background to lower the priority of backup operation processes. To learn more about the process priority setting, see this Microsoft article.

Keep in mind that Veeam Plug-Ins manage the priority of backup operation processes only if the system running on the target computer is busy with other tasks.

To throttle Veeam Plug-In performance, do the following:

1. Start the Veeam Plug-In configuration tool using one of the following commands depending on the OS you are using:

• For Linux or Unix:

• For Microsoft Windows:

2. To confirm that you want Veeam Plug-In to throttle performance, type y.

Moving Digest Calculation to Backup Repository

During backup, Veeam Plug-In calculates a digest, or checksum, for each transferred data block. Veeam Backup & Replication uses digests to verify the integrity of backed-up data. By default, Veeam Plug-In calculates digests on the IBM Db2 server. If throttling does not reduce the CPU load on the IBM Db2 server enough, you can move digest calculation to the backup repository.

Before you move digest calculation to the backup repository, consider the following:

• Moving digest calculation to the backup repository will reduce the CPU load on the IBM Db2 server, but will increase the load on the server with backup repository.
• If you encrypt backups, Veeam Plug-In ignores the Veeam Plug-In configuration file and always calculates digests on the IBM Db2 server.
• Veeam Backup & Replication can calculate digests on uncompressed data only. As a result, if you want to move digest calculation to the backup repository, you must send uncompressed data from the IBM Db2 server to this repository. As a workaround, you can compress data on the repository after the digests are calculated. For details on compressing data on the repository, see step 2.

To move digest calculation to the backup repository, do the following:

1. Open the Veeam Plug-In configuration file with a text editor. The file is located in one of the following locations:

• For Linux or Unix:

/opt/veeam/VeeamPluginforDB2/veeam_config.xml.

• For Microsoft Windows:

%PROGRAMFILES%\Veeam\VeeamPluginforDB2\veeam_config.xml.

2. In the veeam_config.xml file, find the <AgentParams /> line and add the following attributes:

With these attributes, Veeam Plug-In sends uncompressed data to the backup repository. You do not need to update the compression settings using the Veeam Plug-In configuration wizard or tool.

Alternatively, if you want to calculate digests and compress data on the repository, add the following attributes to the Veeam Plug-In configuration file:

In this case, Veeam Plug-In sends uncompressed data to the backup repository. Veeam Backup & Replication calculates digests and only then compresses the data.