Protection plan runtime may be impacted by several conditions, including memory use, server load, priority settings , and the amount of active data scanned.
Plans are initiated by Windows scheduled tasks. The priority of the scheduled task that initiates the protection plan can impact plan runtime duration. In prior versions of the software, the scheduled task were created with 'below normal' priority; new tasks created in the 9.0 release of the software run with 'normal' priority. To check the scheduled task priority of a particular plan, highlight the protection plan, then click the 'edit schedule' action in the right pane. The priority is listed on the Additions tab of the task. If the task is set to run with 'below normal' priority, set the plan's scheduled task to run with normal priority using our kb, Increase CPU priority of protection plan scheduled task.
In addition to the scheduled task priority, you can adjust the protection plan priority. Highlight the protection plan, then click 'Advanced Settings' in the right pane. Set the priority field to 'below normal', 'normal', or 'above normal' as desired. Click OK to save the setting change.
If you confirm that a long plan runtime occurs even though very little active data was encountered by the plan, you may run the plan with boost disabled. Go to Archive Manager > Remote Computers > [the computer] > history and highlight the result with a long runtime. The number of new and changed files may be small, as well as a small amount of changed data in MB or GB. Further the plan log may indicate the plan's individual processes end much earlier than the entire plan. When boost is enabled, a final scan of the file system takes place after individual processes end. In this case, go to the plan's advanced settings and set the boost field to false. Click OK to save the advanced settings change.
When history is reviewed, notice whether the plan has been ending with errors and warnings. Plan indexes may not get trimmed in this case, allowing them to grow. Indexes take system memory so rebuilding them may free up system resources and allow the plan to run more efficiently.
If multiple plans are scheduled to run on the same server, adjust the schedules for each so they do not run concurrently. With boost enabled, each plan will run with multiple processes, increasing the CPU usage, memory usage, and disk IO. The total plan processes will equal that number times the number of plans running concurrently. At some point this may impact plan runtime duration. If plans must run at the same time, then disable boost or throttle the number of processes boost uses. Refer to Performance tuning protection plan settings.
For the Exchange Databases plan type, the plan performs a database integrity check, by default. The integrity check can be tuned for better performance using the throttle field in the plan's advanced settings. Default values for throttling are in effect for different versions of Exchange. See chart:
Plans are initiated by Windows scheduled tasks. The priority of the scheduled task that initiates the protection plan can impact plan runtime duration. In prior versions of the software, the scheduled task were created with 'below normal' priority; new tasks created in the 9.0 release of the software run with 'normal' priority. To check the scheduled task priority of a particular plan, highlight the protection plan, then click the 'edit schedule' action in the right pane. The priority is listed on the Additions tab of the task. If the task is set to run with 'below normal' priority, set the plan's scheduled task to run with normal priority using our kb, Increase CPU priority of protection plan scheduled task.
In addition to the scheduled task priority, you can adjust the protection plan priority. Highlight the protection plan, then click 'Advanced Settings' in the right pane. Set the priority field to 'below normal', 'normal', or 'above normal' as desired. Click OK to save the setting change.
If you confirm that a long plan runtime occurs even though very little active data was encountered by the plan, you may run the plan with boost disabled. Go to Archive Manager > Remote Computers > [the computer] > history and highlight the result with a long runtime. The number of new and changed files may be small, as well as a small amount of changed data in MB or GB. Further the plan log may indicate the plan's individual processes end much earlier than the entire plan. When boost is enabled, a final scan of the file system takes place after individual processes end. In this case, go to the plan's advanced settings and set the boost field to false. Click OK to save the advanced settings change.
When history is reviewed, notice whether the plan has been ending with errors and warnings. Plan indexes may not get trimmed in this case, allowing them to grow. Indexes take system memory so rebuilding them may free up system resources and allow the plan to run more efficiently.
If multiple plans are scheduled to run on the same server, adjust the schedules for each so they do not run concurrently. With boost enabled, each plan will run with multiple processes, increasing the CPU usage, memory usage, and disk IO. The total plan processes will equal that number times the number of plans running concurrently. At some point this may impact plan runtime duration. If plans must run at the same time, then disable boost or throttle the number of processes boost uses. Refer to Performance tuning protection plan settings.
For the Exchange Databases plan type, the plan performs a database integrity check, by default. The integrity check can be tuned for better performance using the throttle field in the plan's advanced settings. Default values for throttling are in effect for different versions of Exchange. See chart:
Exchange 2003 = 500, Exchange 2007 = 125, Exchange 2010, Exchange 2013 (and newer) = 32
For Exchange 2010 and newer versions, a number larger than the default 32 represents a longer pause between checks, so the throttle would decrease the percentage of CPU time used, and have less impact on the server, but the check would take longer to complete. A number smaller than 32 represents a shorter pause between checks, increasing CPU time, and have more impact on the server, but the check would take less time to complete. Be aware that some throttle values may impact the server noticeably.
For Exchange 2010 and newer versions, a number larger than the default 32 represents a longer pause between checks, so the throttle would decrease the percentage of CPU time used, and have less impact on the server, but the check would take longer to complete. A number smaller than 32 represents a shorter pause between checks, increasing CPU time, and have more impact on the server, but the check would take less time to complete. Be aware that some throttle values may impact the server noticeably.