same issue. 10 days before we have upgraded our environment to Nov 2019 due to internal engine error with June 2019 .. but the issue still occuring. By any chance you logged case with Qlik? Any update?
Our Qlik was depolyed in a multi-node environment. The issue was with the CPU and mainly RAM/Memory utilization. It was bettween 90-100%. After increasing the RAM we have resolved this issue. RAM is consumed by 75% whereas, COU fluchuates b/w 75%-100%.
Qlik mian events are handeled by RAM. Increasing it has resolved. Also, the service account should be sepreate from the admin account for HUB and QMC. Generally, this also exhauts the resources making the normal Operation at Qlik's end.