Description of problem: On a system with a LOT of repositories, during migration the sidekiq worker for dynflow will consume ~16GB, which is way too much. We need to look at breaking up the errata import process to not consume as much memory. In this case, there were ~25,000 errata, and 21,153,436 errata <-> repo associations (the biggest issue)
And here was we witnessed in /var/log/foreman/production.log: 2021-03-18T15:07:35 [E|bac|] failed to allocate memory (NoMemoryError) | 2021-03-18T15:07:35 [I|bac|] Task {label: Actions::Pulp3::ContentMigration, id: 91601a75-10d2-47bb-af1e-891c91efeb2e, execution_plan_id: ac7599d1-e114-42ec-9db4-02ae0eadb41d} state changed: stopped result: warning 2021-03-18T15:07:35 [I|bac|] Task {label: Actions::Pulp3::ContentMigration, id: 91601a75-10d2-47bb-af1e-891c91efeb2e, execution_plan_id: ac7599d1-e114-42ec-9db4-02ae0eadb41d} state changed: stopped result: warning There was also a dynflow traceback in /var/log/messages, but that was likely more of a side effect
Created redmine issue https://projects.theforeman.org/issues/32142 from this bug
Verified on 6.9.0 snap 19.1.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Moderate: Satellite 6.9 Release), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2021:1313