with 0 or 1, you should activate the dynamic preemption of PCH by AGCH (called AG_PREEMPT).
The KPIs used to monitor PCH and AGCH resources are:
– PCH blocks discarded due to congestion
– AGCH blocks discarded due to congestion
So my advice, as a starting point :
AG_PREEMPT = Disable
AG_BLKS_RES = 4
Then once you have an idea about your PCH load and AGCH load, you can fine tune it, and maybe activate the dynamic preemption.
BTW, when dynamic preemption is enabled, you can’t know what are the loads of agch and pch, as their capacities is now “dynamic”. It prevents any possible load computation. However the blks discarded due to congestion are still available.
when agch is congested (as it certainly was when you put ag blks res = 1), immediate assignments are discarded. This will increase the agch congestion %, obviously, but not any other KPI…
This should not impact cd% or hos%, so I don’t know what happened in your network.