• jacroberts (5/7/2010)


    Grant Fritchey (5/7/2010)


    Also, we make it a point of retaining the scripts that were generated for the final rollout to production, in case there's any question in the future what it was that got deployed.

    When we release to an environment we 'label' the files in PVCS so at a later date we can get any release that has gone into any environment just by getting the files by the relevant label.

    We cut a branch (we use Perforce) at "code freeze". Contained in that code branch is the application code (rich client, web tier, BLL, database changes, etc.) The schema creation (for new customers/databases) and update (for existing customers/databases) SQL scripts are then used for QA and eventual deployment. We also version stamp the schema in each database. So the upgrade script will fail right out of the gate if it is being applied to a database that is not the correct schema version. ALL changes to the database are done via scripts. So we know exactly what is running in production.


    [font="Arial Narrow"](PHB) I think we should build an SQL database. (Dilbert) What color do you want that database? (PHB) I think mauve has the most RAM.[/font]