You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
On my local Linux machine, running the engine, this takes 32 seconds to render. It's a bit slow, but it gets there.
Problem is, the modeling app stream disconnects after several seconds. Then it tries to reconnect, waits, disconnects, reconnects, again and again. Mike noticed these logs in the console:
Our theory is that when the engine takes a while to respond, the stream disconnects. But it should be more patient. A slow engine response that eventually gets there shouldn't cause this disconnect-reconnect loop. Frontend should tolerate a slow engine better.
@jtran originally noticed this problem tonight, on the 8020 rail. He and @franknoirot debugged it and eventually called in @mlfarrell to help from the engine side. We've established the engine is slow but handling it fine, not crashing or anything. So I think this is a frontend timeout/stream issue.
The text was updated successfully, but these errors were encountered:
I've started to look into this. I've opened up another ticket with some internal details on what I've found. I don't think the modeling app is involved here.
This file renders a 100,000-sided polygon.
On my local Linux machine, running the engine, this takes 32 seconds to render. It's a bit slow, but it gets there.
Problem is, the modeling app stream disconnects after several seconds. Then it tries to reconnect, waits, disconnects, reconnects, again and again. Mike noticed these logs in the console:
Our theory is that when the engine takes a while to respond, the stream disconnects. But it should be more patient. A slow engine response that eventually gets there shouldn't cause this disconnect-reconnect loop. Frontend should tolerate a slow engine better.
@jtran originally noticed this problem tonight, on the 8020 rail. He and @franknoirot debugged it and eventually called in @mlfarrell to help from the engine side. We've established the engine is slow but handling it fine, not crashing or anything. So I think this is a frontend timeout/stream issue.
The text was updated successfully, but these errors were encountered: