General guidelines for planning the update process
A reference topic that contains general guidelines to create a standard process for moving customizations from instance to instance.
-
Use a standard path for update sets: migrate from development to test, then from test to production. Avoid moving the same update set from more than one source.
- Check that both instances are on the same version. While it is possible to move an update set from an instance on an older version to an instance on a newer version, it increases the likelihood of problems. Customizations may not work if they rely on code that has changed between versions.
- Change a single update set for each small or medium task. Large update sets with schema changes or major workflow revisions are tougher to review, slower to process, more conflict-prone, and take longer to preview or commit.
-
Ensure that all base system records have consistent sys_id fields. Certain base system records may be generated on an instance post-provisioning, resulting in differences between instances and potential complications with update sets.
-
Schedule update set commits outside business hours to avoid slowing down the production instance. Any reduced performance is brief.
- Use clear update set names and establish a naming convention to coordinate developer changes and streamline referencing during commits.
- If update sets are being generated as fixes for problems, consider including the problem ticket in the name. For example PR10005 - Duplicate Email Issues Fix.
- If you need more than one update set to address a problem, include a sequence number in the naming convention. Sequenced naming conventions confirm that update sets are applied in the order that they were created. For example, PR10005 - Duplicate Email Issues Fix and PR10005.2 - Duplicate Email Issues Fix.