How does UBI handle an unexpected litigation hold order (“in-place” or “e-discovery” hold)?

In this scenario, you have a node that is already backing up to UBI on a regular schedule, with one of our normal retention policies.  You receive an order regarding upcoming litigation that you must retain all current copies of your backups, or that you must turn over copies of all current backups.

1. Immediately notify the UBI administrators, via our ServiceNow form, that you need to place an e-discovery hold on your node.  Include your node name and UBI server name in the request.

Upon receiving your request, we will take the following actions:

1. We will immediately lock the node to prevent user access like client-initiated deletions.  Once this occurs, you will not be able to perform new backups or restores until the node is unlocked later in the process.
2. We will immediately move the node into our “e-discovery” policy domain for temporary holding.  In this policy domain, all retention policies are configured for unlimited retention (both in duration and version count).

How does UBI handle a planned, ongoing litigation hold order?

In this scenario, you have been told that all future data on your machine, or all future backups from your system, must be retained indefinitely.  This may need to continue for many years before you will be able to resume following normal retention policies.

Spectrum Protect is not designed to handle this scenario well, so it is not a good fit for UBI.  We strongly recommend using a different product that is better equipped to handle this scenario.  Some of the limitations are:

• Backups are client-controlled and periodic, so there is no guarantee that all files or versions are being captured.  Files can change multiple times during the day, but we will only capture the file as it existed during the backup.  Users can inadvertently or intentionally exclude files which should have been preserved, and we will not be able to detect or correct this.
• While the server does know whether a node attempted a backup, it does not know whether the backup skipped files or terminated early, so we cannot know if backups are truly successful.  For example, files might be skipped if they are in use, or if their permissions are set incorrectly.  You must regularly examine your backup logs to ensure that your backups are complete and that you are complying with the litigation hold order.
• Even if your backup logs report that backups are completing successfully, it is unlikely but possible that client bugs or random corruption may cause files to be stored incorrectly.  For example, IC57348 caused 5.x clients to fail to back up correctly while reporting success.
• All file metadata is stored in a central database, which creates a limit on the number of file versions a single UBI server can manage at one time.  Allowing unlimited retention causes the server’s database to steadily fill until eventually the server must be taken out of active service.  This may require that we issue new nodes for users of the server, resulting in fragmentation of litigation hold data across multiple servers and accounts.

When you request a new node via our ServiceNow form, make sure that you attach a copy of your litigation hold order and answer “Yes” to the question “Have you been informed that your data is subject to Litigation Hold?”

You will be placed on the litigation hold server. The server receives backups and archives from nodes, replicates that data off-site, and otherwise behaves, from the user’s perspective, in exactly the same way as all of the other UBI servers.  The retention for these nodes will be set to indefinite.