Description

Running setup.sh (for RPM installs) or the installer playbook (for containerized installs) is a task that always benefits from quicker run times.

Inventory files, now (i.e. on AAP 2.5) with gateway and redis clusters, very often reach

Customers usually run the install script or playbook from a machine beefy enough to handle ~15 concurrent ansible processes.

It would be in our customers' (therefore also our) interest if the installer would automatically adopt a fork count of ~15, perhaps even higher. Installer runs would complete quicker when everything is alright, and would fail quicker when troubleshooting.

This could be accompanied by an addition to the install documentation, stating that running the installer from very small machines would require explicitly setting the fork count to a lower number.

For one, the addition of `-- -f12` to this one support customer's `setup.sh` command resulted in a "screaming" (their words) speed up in setup.sh run times. Their inventory contains 12 hosts. They were completely unaware that specifying a higher fork count was even possible for the installer, and this tidbit helped reduce customer temperature around this particular ticket where they were running setup.sh multiple times for their multiple issues.