CloudM Migrate enables document migrations to run in a dedicated worker process, allowing mail, contacts, calendars, and other data types to be processed in parallel. This significantly accelerates the overall migration process for each entity. This article covers some frequently asked questions related to this feature.
Is this option enabled by default?
Yes, this option is enabled by default. However, you can enable or disable it at the migration batch level.
To adjust this setting, navigate to the configuration step of the migration batch, then go to:
Advanced Settings > System and toggle the "Process data types in parallel" option.
Â
When Should I Disable This Option?
While parallel processing improves migration speed, there are scenarios where you might prefer to disable it.
For example, if a user has a small volume of mail but a large number of documents, you may want their mail migration to complete first before processing documents. This prevents two worker processes from being allocated to a single user, optimizing resource usage.
However, for most cases, it is recommended to keep this option enabled to maximize migration efficiency.
How does this impact the concurrency of migrations?
By default, in Migrate Self-Hosted a secondary service can run up to 20 processes simultaneously. This limit can be adjusted in the CloudM Migrate Service Manager by modifying the "Maximum User Migrations" setting. Unfortunately, this cannot be modified in Migrate Hosted.
Typically, when migrating a batch of users, each user consumes one process, allowing up to 20 users to be migrated concurrently.
With parallel data types enabled, each user utilizes two worker processes—one for document migration and another for all other selected data types. As a result, only 10 users can run concurrently, but the overall migration efficiency improves, enabling users to complete their migrations faster.
The migration service always utilises an even number of worker processes when processing data types in parallel. If the maximum user migrations setting is an odd number, the service will exceed this limit by 1 to maintain an even worker count.
Example:
- If the setting is 5 and 3 users are processed in parallel, the service will allocate 6 worker processes instead of 5.
This ensures optimal resource distribution and parallel execution efficiency.
Is this option available in Migrate Hosted?
Yes, this option is available in both Migrate Hosted and Migrate Self-Hosted, and it is enabled by default in both.
How do I know what users are processing in Parallel?
During the migration, in the overall migration statistics, you will see a line indicating how many active migration processes are running as well as how many users are processing in parallel.
Additionally, you will now be able to expand the row for each user processing in parallel to get a breakdown of the number of documents migrated as well as the number of other data types migrated, depending on what has been selected for migration.
The below screenshot shows an example of a user processing in parallel.
Can I stop and start individual processes?
Yes, you can stop or start individual processes or actions for the entire user. Simply hover over the associated action icon and select the desired action.
What impact does this have on trace files?
Trace files are generated for each process. If a user is running processes in parallel, separate trace files will be created for documents and other data types.
Does this work for other entities?
Yes, this works for users as well as any other entity that includes documents and other data types.