Description of problem:

After creating an ACS on the Satellite, the task runs and never progresses after 60%. Looking at the system utilization, all memory and swap are being used on the Satellite rendering the Satellite unusable.

How reproducible:

every time

Is this issue a regression from an earlier version:

Not to my knowledge

Steps to Reproduce:

1. Install Satellite 6.15.1.1 with 4 CPUs and 20GB RAM
2. Attach manifest with SKUs RH00798 and RH00270
3. Enable repositories rhel-6-server-rpms and rhel-6-server-els-rpms and sync
4. Create a Simplified Yum ACS with the Satellite selected for the capsule, no http proxy, and all repositories added (just the 2 RHEL 6 ones enabled previously).
5. Refresh the ACS after creating it and wait for the resources to all be consumed (took ~2 hours for me)

Actual behavior:
2 pulpworkers are consuming 43 and 28% of all the available memory on the Satellite. All RAM and SWAP are completely utilized.

Expected behavior:
Satellite should not be rendered unusable by this task

Business Impact / Additional info:

No other task was running on the Satellite during this time. This task has currently been running for 4 hours so far with no further advancement on status.

foreman=# select * from foreman_tasks_tasks where state = 'running';
[ RECORD 1 ]--+-------------------------------------------------
id               | 7acffdd0-6170-4244-8da9-fac9ee1a723c
type             | ForemanTasks::Task::DynflowTask
label            | Actions::Katello::AlternateContentSource::Refresh
started_at       | 2024-07-01 14:31:13.351
ended_at         | 
state            | running
result           | pending
external_id      | 038da65c-7650-4f47-ba5e-964e9baaa6fe
parent_task_id   | 
start_at         | 2024-07-01 14:31:13.322497
start_before     | 
action           | Refresh Alternate Content Source
user_id          | 4
state_updated_at | 2024-07-01 14:31:14.662898