Hey working on a big data migration I've stumbled across this project. Here is my current situation:
OnPremise application with a filestorage of 120GB with a lot of small files (500k files)
Filestorage needs to be migrated to Azure Files with the only possible way of mounting the Share is SMB
All data must be copied
Downtime of the application must be as short as possible.
cp is reasonable fast only a couple of hours but rsyncing afterwards is not an option.
My current approach would be now:
Create migratefs (upperdir=/mnt/azure lowerdir=/local/file/storage) and relink application
cp -ru /local/file/storage /mnt/azure
Should I better use /mnt/migratefs as copy target to avoid problems with write operations?
Is there an option to mimic this behaviour? If so, how do I know that is is done?
Hey working on a big data migration I've stumbled across this project. Here is my current situation:
OnPremise application with a filestorage of 120GB with a lot of small files (500k files) Filestorage needs to be migrated to Azure Files with the only possible way of mounting the Share is SMB All data must be copied Downtime of the application must be as short as possible. cp is reasonable fast only a couple of hours but rsyncing afterwards is not an option.
My current approach would be now:
Should I better use /mnt/migratefs as copy target to avoid problems with write operations?
Is there an option to mimic this behaviour? If so, how do I know that is is done?
Thank you already so much for this project.
Best regards
Flo