Dear HWS Community @everyone
unfortunately we see those disconnects and issues (especially HWS NA) you have, with big worry. We are not 100% sure why yet and keep investigating but few points:
- it COULD be that it’s because of the AMD CPU on HWS NA. We wanted to give AMD a chance but maybe Intel is still the superior here.
Reminder about our Hardware:
Server Hardware and Cost
-
the galaxy is not the enemy here. HWS 11 had over 200 playfields too and we had the same or little less opened playfields at the same time. I crawled the database and found many playfields with less than 10 frames per second (FPS). THIS is why you disconnect so often.
It’s called “playfield process weight”.
And for that a change came in with Alpha 12. In HWS 11 PvP playfields claimed a whole process for their own. It was not shared with other playfields. Now it is shared again. So it could be that ECC and Golden Globe could share their weight…
There is a new flag in the playfield.yaml to control the process weight per playfield.
I will try it out on HWS NA tonight. For now only Golden Globe.
Explanation: if multiple playfields share their process it is in general good for hardware performance. Less process = less hardware consumption. But sharing… well, if one gets broken, others too. -
As a temporary test… we also set the player slot count of HWS NA tonight to 130. Maybe it will show us if it’s because of the CPU or not.
Sorry overall for the inconveniences. I’m quite upset about it.