Deploy container image on AWS Lambda - residual timeout : why?

[will-afs]

• Not enough RAM?
• Not enough storage?
• Not enough network IO?
• Can not access tmp filesystem?
• Permissions problem?

[will-afs]

Not enough storage: 512Mb seems ok

Not enough RAM: tried with 3008 Mb so it seems ok

Network IO cannot be configured

[will-afs]

Ok from container, but not ok when deployed on Lambda.

Seems to have a problem with refextract :

• contacting the internet from AWS Lambda
• or creating a temporary file on /tmp

[will-afs]

==> Testing the file creation when deployed ==> Created

==> Testing timeout in applicative code ==> No problem
==> Testing timeout in deployed container ==> triggered

==> Problem comes from Lambda access to external world (is Lambda authorized to request outside world?)

[will-afs]

Solution n°1 : connect the VPC to external world ==> https://aws.amazon.com/fr/premiumsupport/knowledge-center/internet-access-lambda-function/

Solution n°2 : avoid any dependency with external world ==> uploading pdf file through request instead of providing its URI

[will-afs]

Problem was that Lambda could not reach Internet because was inside a VPC not connected to it.

Solved the problem by following this tutorial: https://www.youtube.com/watch?v=JcRKdEP94jM