If you delete a user in CentreStack, with the default settings the system will remove the user from the CentreStack database but will leave all of the user's personal files/folders intact in the backend storage of the tenant. This article may help you if you would like to restore the deleted user's backend storage folder from outside of the CentreStack application. Most user's root files will likely not need any special treatment, but if a user has content that was synced (attached, migrated, or linked) from a windows client, mac client, or server agent, then the backend folders (gsync folders) will be appended with IDs, timestamps, action markers, and will be very difficult to restore manually.
There is one setting under the Cluster Control Panel->Worker Nodes->Cloud Monitor Settings that you can enable so you get an option before you attempt to delete a user. If you say yes to this checkbox option , the user's storage will be purged by the Gladinet Cloud Monitor on the CentreStack server. But, since this is not the default setting, you should be able to restore a deleted user's files as long as you don't have DARE (data encryption at rest) enabled on the tenant, and the storage has not been purged.
For a thorough explanation of what the backend storage looks like, please take a look at this article:
Please be aware that this solution is not supported and is recommended only if you have no other options/backups. Having said that, if you are ready to restore a gsync folder (a cloud copy of a folder that was attached to CentreStack) back to it's original human-readable state despite the obvious risks inherent with batch renaming all the files in a folder, you can try the script included on this page at your own risk.
The way the script works is by doing the following tasks in order:
- All folder deletion markers will be deleted
- All head versions (most recent versions) of all files will be renamed back to their original names
- All non-head versions of all files will be renamed with a common "_$DELETE$" marker (in order to make searching easier)
- All files with a deletion marker will also get the common "_$DELETE$" marker
In order to run it, please do the following:
Make a backup copy of the user's backend storage folder somewhere on another computer (please do not run this script directly on the production environment). If you have trouble copying from the backend storage because the file paths are too long, then use the windows's built-in robocopy command which gets around this issue (example: robocopy "[source]" "[destination]" /e).
- Run the exe file. Supply the full path to the location of the root folder that you want to restore
- Press enter to continue. The script will run recursively and rename all the files back to their original names. Version files will be renamed with a common marker so you can easily search and batch delete them all at once.
- Now, actually delete the files by performing a search on the entire folder with windows explorer for the string: "_$DELETE$"
- Select all marked files (control+a) and press the delete key on your keyboard
Here is the script!
Enter the path to the folder to be restored:
After running the script, all files that are OK to delete will be easily spotted (with _$DELETE$):
Final result (after deleting the marked files):
Please submit a ticket to firstname.lastname@example.org if you have any issues.