We believe we are running into GVL lock (with some intensive CPU requests) but can’t really prove/disprove. If support for gvltools and a visual indication per request of how much time it was ready but waiting would be really good
We believe we are running into GVL lock (with some intensive CPU requests) but can’t really prove/disprove. If support for gvltools and a visual indication per request of how much time it was ready but waiting would be really good