Open
Description
Results of the migration:
- List of the migration by user:
For each migration, we need to have:- Number of the migrated files
- Total number of files for the user
- Number of the migrated files
- Number of errors
For every error, we need to have the type of error(if we can determine)- stacktrace
- input data that caused an error (user, doc id, file id, etc)
List of the errors that we know and need to deal with:
- DNS resolution of the user's hostname
- Unknown errors with the empty response from the cozy stack
- Files don't exist in S3
- Metadata is corrupted(undefined 'id' field)
To run several migrations in parallel, we need to split the user list into batches.
To be simple and not use a distributed lock on the user, we can continue to consume users from command params, but we need to pass a lot of emails as well as it's not so comfortable to do with params, so let's also introduce env variable for the users list for the migrations.
Metadata
Metadata
Assignees
Labels
No labels