its something with resets + streaming +
# spicedb
j
its something with resets + streaming + gRPC node lib
t
I see. I've been digging these issues and it seems that the only known workarounds are downgrading grpc-node to v1.7.3 (which would require me to fork the authzed node cliet), or disabling retries...
y
yeah, that's also what we've seen :/
t
How bad would you say it is to disable the client retries?
y
not that bad, especially if you write some application-level logic to retry on those errors
t
I tried setting
{ 'grpc.enable_retries': 0 }
on the client sdk, but the error still kept happening