i have application block block (x pixels x pixels) processing (which can paralleled) in python. need advice on best approach is, architecturally.
use lambda function when image uploaded s3, processing (however, i'm confused how can take advantage of parallelization or hpc, invoking multiple lambda functions each part of image being processed?)
using sqs , ec2 (i'm doing research on this, hitting wall).
i wanted know better approach be. continually researching, , realize time valuable, wanted see if here had done similar.
thank you!
i'm don't think there 1 single best way dependent on project, goals, constraints etc.
as example, i've worked on similar sounding problem: had process svg file depending on data provided lambda function.
initially looked @ doing similar (1) s3->lambda didn't fit problem domain because of nature of lambda function being triggered s3 object-created event. better fit me able call lambda function on demand.
due ended calling lambda function directly via http request api gateway (api gateway->lambda). potentially allow call different lambda functions per part of image being processed again dependent on trying achieve.
Comments
Post a Comment