BEAR Archive

 BEAR Archive@640px

Archiving data associated with Research Projects

Note that the BEAR Archive is not a service for the storage of data associated with publications for the purposes of satisfying open data requirements. Instead, you should use UBIRA eData, the University of Birmingham Institutional Research Archive.

The purpose of the BEAR Archive is to store valuable data associated with your research project (current or completed) which may be needed in the future but which is not currently being actively used and is not needed to support a publication.

Data stored in the BEAR Archive will be archived to tape and retained for 10 years. The data will only be accessible upon an application for restoration by an appropriate person (e.g. the PI). After this period the data will be permanently deleted automatically, with no notification.

Types of BEAR Archive

There are three scenarios for using the BEAR Archive. Please see "How to Apply for the Archive" below for instructions on how the project’s Principal Investigator (PI) or their registered delegate can do this. All these scenarios involve archiving the contents of a Research Data Store (RDS) project.

  • End of project: When a research project has ended and access to the data in the RDS is no longer required, then the contents of the RDS project will be archived to the BEAR Archive.
  • Project registration expired: All RDS projects need to be re-registered regularly. If a project has not been re-registered for a long time (usually over 2 or 3 years), then a service desk ticket is automatically created informing the PI that the project will be archived to the BEAR Archive. The PI or delegate can respond and re-register the project to keep it active, and the archive process is cancelled.
  • Mid-life archive: A PI or delegate can request a mid-life archive of an RDS project. In this case a new special kind of project is created and linked to the original project. The PI is then responsible for moving the data from the main RDS project into this new project, after which the new project will be archived. An example of a mid-life archive might be if the RDS project is nearing its quota, and contains a substantial dataset that is no longer needed for active use and so can be archived, but might be needed again in the next 10 years. Archiving it will free up space in the active RDS project.

How to Apply for the Archive

Applications for storage in the BEAR Archive should be made by the project’s Principal Investigator (PI) or their registered delegate, using the IT Service Desk ( 

  1. Complete the 'Archive a BEAR Project form' for each BEAR project to be archived, specifying the BEAR project code (e.g. "smithaa-abc-xyz")
  2. You will be required to confirm that you have read and understood the policy detailed in the form, you can then submit the form. 
  3. The system will then create a Service Desk ticket on your behalf to archive your project as an "end-of-project archive". If you want a mid-life archive (see "Types of BEAR Archive" above) then please add a note to the ticket explaining this and please provide any additional information you think will be useful.

This process will archive any BEAR Cloud Virtual Machines and GitLab groups associated with the project and put the data in the project's RDS storage. It will then archive the RDS storage to tape.

Note that as an archive solution, the data will be removed from the RDS and access to the data in future will require a request via the IT Service Desk. As the solution is intended for cold storage of data, recalling data can take a significant amount of time.

Archive Costs and Funding

The University funds a reasonable allocation (currently 20TB) for the archive requirements of research projects for 10 years. However, IT Services reserves the right to charge for exceptional requirements i.e. more than 20TB. Funding for exceptional archive needs should be directly included in the research bid.

Please note that the total archive space needs for a project includes any mid-life archives - and so if multiple mid-life archives add up to more than 20TB then this will need to be paid for.

Sample text for a grant application is given below:

'The University provides a standard archive storage allocation for the retention of data associated with the project, however if this project is likely to generate data in excess of the 20TB allocation there will be a charge. The cost charged by the University is currently £40 per additional TB for up to 10 years storage.'

The Technical Solution

The solution uses an enterprise class ‘data protection platform’ (software) and tape storage media. It has been created to be fault tolerant, with data duplicated between two data centres to minimise the risk of data loss. Tapes are periodically and automatically verified to help maintain the integrity of the data. All tape media is encrypted as data is written to the tape. The use of tape media provides a low-cost storage medium for long term retention of data.

Data is automatically expired from the archive once the 10 years has passed, with no warning or notification.