Performance Questions

[drfraker]

Hi. First of all, thanks for the code. I appreciate your work on this project. I have this running on aws Lambda and everything seems to be working great except for 2 things.

1. I'm seeing 2.5 second processing with very basic 11K PDF's.
2. If the PDF contains links to large images (2 images ~4M each in my case) the PDF will fail to be generated and throws an error.

My questions are:
Is this the normal processing speed? Is there anything I can do to improve the processing speed?
Do you have any idea why the PDF fails to generate with large images?

[elverado]

Just getting into using this myself, but my past Lambda experience tells me that your 2.5 second processing may be the Warmup penalty inherent in Lambda; do immediately subsequent attempts run any faster?

As for #2 I'd be curious what your memory and CPU settings are for your lambda container config. That and any limits to the content-length of your HTTP plumbing (gateways, client code, etc). The heart of this being wkhtmltopdf it's really spinning up a full web browser rendering engine QT Webkit behind the scenes and running a 'print to pdf' on it. So it's understandable it's a memory hog and a bit slow. But one of the best options for CSS rendering, at least on the non-paid library side of things.