Support serial loads with update_data task
JonnyPower opened this issue · comments
Feature Type
I'm interested in functionality that solves a current problem or limitation in how CumulusCI works.
Does your feature request solve a current problem?
When running update_data, cci may create a bulk data job in Salesforce, by default taking advantage of parallel jobs via the bulk job's concurrency setting.
However, there are use cases that require serial job runs to prevent e.g. record locking issues.
We intend to leverage lots of update_data tasks as part of our ISV package's upgrade automation, but in some cases running with a parallel concurrency mode will require retries / hurt task stability during our automation.
Describe the solution you'd like
As part of the update_data task, either load a similar .load.yml file as we can define elsewhere to set the concurrency mode, or a cli flag to set api options defined as an empty dict here: https://github.com/SFDO-Tooling/CumulusCI/blob/8e6fa8aeb74ba184a7db0cd589c7c2f791318f71/cumulusci/tasks/bulkdata/update_data.py#L211C20-L211C20
Describe alternatives you've considered
No response
Additional context
We'd be happy to help implement this feature if you have a recommended / preferred approach to solve