AWS node lambda bug

Hi, since few weeks sentry stop auto-catch errors from code on my AWS lambda platform. Manual captureException function always works like a charm. Instead of unhandle errors were logged normally in sentry when I run code in local, log in sentry failed once deployed on AWS lambda. Is this could be due to migration 8.10 on AWS lambda?

sometime I have this message

raven@2.6.3 alert: failed to send exception to sentry: socket hang up
raven@2.6.3 alert: failed to send exception to sentry: Connection timed out on request to sentry.io

Sentry’s latest Node integration on Lambda now support automatically capturing errors without having to call captureException.