Skip to main content

 

Druva Documentation

How user migration in a storage pool works

inSync Private Cloud Editions: File:/tick.png Elite File:/tick.png Enterprise

How user migration works

A storage pool is a collection of storage that function as a unit for backing up data. The storage pool acts as an automatic load balancer and ensures that data is evenly distributed across storage that are part of the pool. You can have unlimited number of primary storage in a storage pool and you can add more primary storage to an existing pool, at any time.

A storage pool with at least one overused and one underused storage is called an unbalanced storage pool.

  • A storage with less than 20% of the total allocated space is an overused storage.
  • A storage with more than 30% space remaining is an underused storage.

When the inSync Master identifies an unbalanced storage pool, it automatically triggers user and data migration from an overused storage to an underused storage. The migration happens without any intervention from inSync administrators. The migration automatically stops when the overused storage becomes relatively free.

Scenarios of user and data migration

Storage A and Storage B with a capacity of 10 TB each is part of Storage Pool Sales. Out of 10 TB of storage space, Storage A utilized 8.2 TB and Storage B utilized 6.5 TB. The inSync Master identifies Storage Pool Sales as an unbalanced storage because Storage A is overused. Thus, the inSync Master starts user migration for the Storage Pool Sales.

The following table lists the scenarios of user and data migration.

Scenario 1:

Storage A: Capacity=10 TB, Utilized=8.2 TB
Storage B: Capacity=10 TB, Utilized=6.5 TB

On Storage A, the inSync Master identifies 4 users with 0.1 TB data size for migration to Storage B.

Result: After migration, the storage consumption for Storage A decreases to 7.8 TB because 0.4 TB of data is migrated to Storage B. On Storage B, the storage consumption increases to 6.9 TB because 0.4 TB of data is migrated from Storage A.

Scenario 2:

Storage A: Capacity=10 TB, Utilized=8.2 TB
Storage B: Capacity=10 TB, Utilized=6.5 TB

On Storage A, all users have 1 TB data size.

On Storage A, all users have 1 TB data size.

Result: If any user is migrated to Storage B, its storage consumption size increases to 7.5 TB and then Storage B is at risk of becoming overused. Therefore, no user is migrated to Storage B.

User migration workflow

The following table describes the user migration from an overused to an underused storage.

Step Description
Step 1 The inSync Master identifies an unbalanced pool.
Step 2 The inSync Master identifies a user for user migration. If any error occurs during user selection, the inSync Master skips user and identifies another user for migration.
Step 3

The inSync Master temporarily stops compaction for the identified user.

Step 4
  • The inSync Master starts data migration for the identified user.
  • The inSync Master temporarily stops the backup and sync operations for the identified user during final stage of data migration.
Step 5 After the data is migrated to the destination storage, the inSync Master checks for errors on the destination storage.
  • If the inSync Master does not find an error, then the inSync Master:
  • Changes storage for the user in the configuration database.
  • Removes data from the source storage.
  • Starts compaction and backup and sync operations.
  • If the inSync Master finds an error, then the inSync Master:
  • Removes data from the destination storage.
  • Skips that user.
  • Identifies a new user.

Note: If the storage pool continues to remain in an unbalanced state, the inSync Master identifies more users for migration.

  • Was this article helpful?