Overview

The purpose of this document is to provide advice on how Flow Scan performance can be increased in certain workflows and circumstances by making configuration changes/choices on the system.

Prerequisites

A Flow installation with some knowledge of the workflow in use and the Flow software

Process Steps

• Don't scan media spaces which you don't need to use within Flow
  FLOW can of course be used across your entire storage ecosystem, but if you have media spaces or storage spaces do not contain useful media (eg temp or render files), these spaces should not be scanned.
  
  Where media spaces contain a combination of useful material and material that does not need to be scanned, those undesirable files can be ignored using the "Ignore List" for each media space in FLOW Control
  
  
• Disable metadata mapping if not needed
  If specific metadata mappings are not required for a particular production, they should be disabled. Even where no mappings are expected to succeed the system will still have to perform extra checks to confirm this.
  
  
• Limit the scope of metadata mapping if needed by applying the mapping to specific media spaces rather than all spaces
  The default storage matching rule is to attempt to map metadata against all space (*)
  Where spaces are not expected to contain material which will produce a match, that matching rule should be restricted.
  For example:
  "LSI*" will match all spaces beginning LSI
  "*media" will match all spaces ending in media
  "rushes1,rushes2,rushes3" will match those explicitly named spaces
  
  
• Configure FLOW automation jobs to scan specific subfolders
  From 2021.3.2, FLOW Automation has the option in the scan task to scan named folders within the media spaces. This can of course be attached to a timed trigger just like one initiated from Flow Control (or any other trigger), targeting your scan more accurately.
• Add the flow scan worker role to additional servers
  From 2021.4 FLOW will have a new role, that of scan worker. This allows the scan role  to delegate jobs to scan workers to work on scans concurrently. To take advantage of this you may need to break monolithic scan jobs with many media spaces (Storage spaces) into separate scan jobs to be delivered to multiple scan workers

Summary

In summary the above steps provide some guidance as to how to improve the performance of flow scans, though not options all will apply in all situations