KohaCon12: Koha Migration Process

by Joy Nelson on Jun 5, 2012

Robin Sheat from Catalyst IT in New Zealand reviewed the process used by Catalyst when migrating libraries to Koha. Migrations of data are detail oriented, repetitive processes. It is best to start this process with a strong plan. The plan needs to contain a timeline outlining what is coming and when the migration is due to conclude. The length of a migration depends on the amount of customization required , OPAC, code and even custom data scripting. The migration plan also communicates what is expected from the migration staff and the libraries. It is important for libraries to test because fixing after a go live can be difficult.

The Basic steps in any process are:

1- Preparation , What software are you coming from and how does the data come out? Is it mrc format, csv or other? CSV files can tend to be difficult to work with depending on the data. Try to use other characters to delimit the field other than ” or , characters.

Identify how items are exported , in marc or separate file? Can you export circulation, borrowers, fines, vendors, authorities, etc. Establish your priority for data and work on exporting the data. This is also a good point to decide on the branches, item types and shelving locations to be used in Koha.

2- First Draft – This phase takes the most time where you will need to create mapping for all fields. The testing begins here and communicating with migration team begins in earnest. Mapping changes and data tweaking are common here.

3, Second draft , A second draft can be implemented to test the migration process. This can be a dry run before the final go live migration. Errors found here are generally easy to address and involve minor tweaking of the migration scripts or processes

4, Go live , This is best done after a library closes for the evening or over a weekend. Errors here can be difficult to fix as the data is in a production environment and is a moving target.