HOW TO PROCEED TO IMPLEMENT SPECIFIC WORKFLOWS WHEN DEALING WITH MULTI TAPE COPIES ?
In this following blog, we are looking to describe 2 additional workflows which can be useful in your day to day data management operations.
- Workflow to manage two copies on tape of the same content
- Externalize one copy of your tape offline and eventually yield the copy to a third party
1- I would like to have all data in a repository residing on two copies so that one tape is allowed to break.
- The first way to achieve this is to create 2 different Data Protection workflows. This will bring the full confidence that the two copies are separated copy process ; and if eventually of the copy process failed, the other one will still be fine.
Dedicate for each workflow a pool "offline archive" and associate for each pool tapes or let Nodeum to get tape from the scratch pool.
For the 2 Workflows : Select Data protection From Source (primary storage : S3 or NAS) to Destination (Tape).
- The second way, if your contents are in a container. You can protect your container on tapes. You have to dedicate one Data Protection workflow to copy the Container data which are still in the cache to a pool "offline archive" and associate for each pool tapes or let Nodeum to get tape from the scratch pool.
The setting of this workflow is : Select Data Protection From Source (container) to Destination (Tape). Dedicate for each workflow a pool "offline archive" and associate for a set of tapes in each pool or let Nodeum get tape automatically from the scratch pool.
When this is executed, then you will have one copy of the Container contents into a set of Tapes.
Afterward, you define a second workflow "data archiving" to archive data of your container to an Active Archive Pool of tape(s). The contents will stay accessible from the Container even if the contents are stored into a pool of the tape(s).
The setting of this workflow is : Select Data Archiving From Source (container) to Destination (Tape). Dedicate for each workflow a pool "active archive" and associate for a set of tapes in each pool or let Nodeum get tape automatically from the scratch pool
2- I would like to have all data of a specific repository on two copies and one of these two I would like to remove from the library and put it somewhere else.
- Either your need is to keep the information into Nodeum to have the possibility in the future to retrieve data from this tape. In this case put the selected tapes out of the tape library. Nodeum understands that the tapes are out the library and Nodeum keep the information into the catalog and report. You have always the possibility to retrieve the data from the tapes which are "out". If you schedule an retrieval operation from one of these tapes, Nodeum will also notify that tapes have to be back into the library before executing the workflow.
- either you want to forget this information. In this case you have the possibility to forget a tape. This is only possible when the tapes are out of the tape library. You can execute this operation from the GUI : Trend Analytic - Tape Reporting and Choose the tape you want Nodeum Forget, then click on the three bullets at the right and select Forget in the list. When this operation is done, the system will removes all data related to this tape from Nodeum (catalog and indexation). Of course the contents which are stored in the tapes, will stay.
3- I would like to copy some data to tape and remove these tapes completely from the library and from Nodeum index, to give it to a third party (e.g; scientist, researcher...).After the departure of a scientist, researcher..., you need to :
- For the 1st set of tapes, you have to remove all data stored in this set of tapes. To remove all data from the tape you don't want to use - Format the tapes and delete the pool. And then do not forget to clean your configuration in removing associated workflows.
- For the 2nd set of Tapes, you give the tapes to the third party, and in this case you forget the tape like mention above (pt2).