#Random crashes

4 messages · Page 1 of 1 (latest)

stark thistle
#

Hi!
We have had some requests reaching our API from which we have no logs (we have built our own logging).

We know it crashes between a call to an external API and a DO request. Because we reach the external API but there is no data in the DO. This whole block is surrounded with a ".catch()" and should have been caught but wasn't.

Is there any possibility that workers can crash in a way that isn't caught by a try-catch/.catch()? And if so how do we handle that?

hexed geyser
stark thistle
#

I'll look into implementing #logs-engine.
But for now is there anyway you know of that can cause a worker to crash in a way which is "uncatchable"?

hexed geyser
#

Returning an invalid response or having a never resolving promise might not be catchable errors that cause exceptions but like I said I would tail the worker logs to see what the error is