With the recent performance push, it's become clear to me that plugin writers probably need to think in batches as a first class concept. If they do then the auto-throttling in pulpcore seems out of place. From a high level, I'm concerned we're accidentally ending up with multiple ways of solving the same problem.
I filed https://pulp.plan.io/issues/5855 to discuss remove auto-throttling from the plugin API in 3.1. Please send feedback with your suggestions on how to resolve the concerns listed. Thanks! Brian
_______________________________________________ Pulp-dev mailing list Pulp-dev@redhat.com https://www.redhat.com/mailman/listinfo/pulp-dev
