I maintain a project that lists hundreds of thousands of web pages, and I’d like to show a screenshot for each web page. There are already some services out there that give you this service, and are very easy to integrate in an app - all you need is use an IMG tag, basically, but I’m worried about the costs.
So, I’m evaluating building this in-house to reduce costs. I think that it would be pretty easy to set up an AWS Lambda with Puppetteer that, given the URL of the page and some options like size, takes a screenshot and maybe caches it to S3?
Do you have any experience with this? Is there already an open source solution that I can easily set up?
I would think it’d be significantly cheaper to create and run you own than use AWS Lambda
You could use something like:
…and host the service on a very cheap dedicated server (you can get them for about $40 a month now I think). You could them simply copy them over to your app or hotlink to them from a sister-domain or subdomain on that server. Large hard drives are very cheap now, so you could easily host up to 2TB for around that cost I reckon.
