File Indexing Process

File indexing works directly with backup data, without connecting to the hypervisor or examining VM group configuration.

A file indexing job includes the following stages:

1. The MediaAgent that is assigned for the backup plan discovers the VMs/instances that are backed up.

2. The indexing job mounts each backup on an access node.

   If auto-scaling is configured in your environment, the Commvault software uses those settings to auto-scale access nodes for file indexing. If the backup process created auto-scaled access nodes, the file indexing process uses those access nodes. Otherwise, the file indexing process creates auto-scaled access nodes.

   The number auto-scaled access nodes that are created is based on the number of CPUs for each access node and the number of streams for each CPU.

   The operating system of auto-scaled access nodes for file indexing matches the OS of the VMs/instances that are indexed.

3. The indexing job checks the region of the VMs/instances to determine whether settings for the same region exist in the associated VM provisioning settings. If yes, then auto-scaled access nodes are deployed in that region. Otherwise, the file indexing job goes into the pending state and the job pending reason (JPR) suggests adding the missing regions to the VM provisioning settings.

4. The indexing job launches a Filescan operation and submits a list of volumes.

5. The Filescan operation processes each volume to create a list of files and folders with the following attributes:

   • Disks and volumes

   • File and folder names

   • File size

   • File type