Description of problem:
During a LEAPP upgrade of a Satellite client, or during its convert2RHEL procedure, the client can easily hit timeout errors when downloading packages from Satellite.

This happens when the repositories with the packages have On Demand download policy (which is the default) and when majority of the packages are not yet downloaded. The feature is not tuned or scaled up for a bulk request of hundreds or even thousands of packages to download from Satellite.

This ends up in repeated upgrade (or convert) failures, always on timeouts of different packages download.

I dont know if this is a known limitation of On Demand policy or a scalability bug. If the later, please fix it. If the first, please reassign the bug to Documentation, to mention in:

https://access.redhat.com/documentation/en-us/red_hat_satellite/6.12/html-single/managing_hosts/index#Upgrading_Hosts_From_RHEL7_to_RHEL8_managing-hosts

and

https://access.redhat.com/documentation/en-us/red_hat_satellite/6.12/html-single/managing_hosts/index#converting-a-host-to-rhel_managing-hosts

that all the repositories the client will fetch packages from are recommended to have Immediate download policy and the repositories should have been synced with that policy, already.

See https://access.redhat.com/solutions/6989411 for a reference of underlying customer problem.

Version-Release number of selected component (if applicable):
Sat 6 (any version incl. 6.12)

How reproducible:
100%

Steps to Reproduce:
1. Sync RHEL8 repos (with the default On Deamnd download policy) to a Satellite
2. Have a RHEL7 client system and run LEAPP upgrade on it
3. Optionally, convert a CentOS / Oracle Linux system to RHEL

Actual results:
2. or 3. is supposed to fail with errors like in https://access.redhat.com/solutions/6989411

Expected results:
Either add to our documentation the recommendation "use immediate download policy", or make the On Demand more scalable.

Additional info: