sls invoke times out before lambda does (2 minutes but lambda is 15)

[scamden]

This is a Bug Report

Description

• What went wrong?
  sls invoke times out after 2 minutes and retries

• What did you expect should have happened?
  invoke waits for the actual lambda invoke to timeout which can be as long as 15 min

• What was the config you used?
  any

• What stacktrace or error message from your provider did you see?
  none

Similar or dependent issues:
#5374

Additional Data

• Serverless Framework Version you're using:
  1.41.x
• Operating System:
  any

[almartin82]

+1, seeing the same timeout.

Error message is

Serverless: Recoverable error occurred (Connection timed out after 120000ms), sleeping for 5 seconds. Try 1 of 4..

Lambda function is set to timeout after 15 min but sls invoke times out at 2 min.

[almartin82]

Looks like this is solvable by modifying AWS_CLIENT_TIMEOUT.

eg:
export AWS_CLIENT_TIMEOUT=90000
sls invoke -f whatever --log

will change the sls timeout. See #490 and #937 for more discussion.

[Pipeliner]

Shouldn't this be set automatically if a longer timeout is configured for lambda in serverless.yml?