Azure blob storage backups use change feed to reduce scan time for incremental backups. This is especially useful for storage accounts that contain millions of objects.
Change feed provides an ordered, append-only log of changes to blobs and blob metadata. Commvault uses this log to identify changes instead of scanning the entire container during incremental backups.
Requirements
Supported storage account types
Change feed is supported on the following storage account types:
-
Standard general-purpose v2 accounts
-
Premium block blob accounts
-
Blob storage accounts
Unsupported storage account types
Change feed is not supported on the following storage account types:
-
Accounts with hierarchical namespace enabled
-
Premium general-purpose v2 accounts
-
General-purpose v1 accounts (you can upgrade these to general-purpose v2)
Permissions
To configure and use change feed, you must assign specific Azure permissions at the subscription level.
Standard roles
-
Storage Blob Data Owner
-
Reader Role
-
Storage Account Contributor
How change feed integration works
-
During the first full backup of a newly created Azure blob instance, Commvault enables change feed on the storage account with a retention period of 90 days.
-
A
$blobchangefeedcontainer is created to store change events for the storage account. Changes to blobs and blob metadata are recorded as.avrofiles in this container. -
Subsequent incremental backups read the change feed data from the
$blobchangefeedcontainer to identify changes since the last backup. This avoids full container scans and improves performance.
Planning considerations
-
You do not need to manually enable change feed. Commvault enables it automatically during the first full backup.
-
Change feed tracks changes to blobs and blob metadata only. Changes at the container level are not tracked.
Frequently asked questions
Does change feed capture blob versioning?
No. Change feed does not track blob versions. Only the latest version of each object is included in backups.