Closed j-adamczyk closed 2 years ago
I understand now what is happening, this is exactly this issue, i.e. I receive logs on shutdown, so for logs for a single event I have to wait until underlying container is killed.
Therefore I change my question: is there any way to force logs flush / another event apart from INVOKE and SHUTDOWN that I can listen to?
I finally fixed this with sleep(5)
, like in this issue. Ugly, weird, but it works.
I tried using this extension to capture logs in Python. The problem is, when Lambda lasts less than a second, I only get messages from extension after about 5 minutes. This renders this essentially unusable, as this is nowhere near real-time logging. Is this normal? How can I fix this?