Site Caching

[Omertron]

Original issue 1680 created by Omertron on 2010-10-27T14:20:53.000Z:

I would be more than happy to give a lot of drive space to caching the data that yamj is getting from internet sites to reuse...

My thoughts:

• skip images.
• different age lengths for movie vs tv show
• don't use cache on a recheck for unknown
• maybe using some of the api has it been updated options like thetvdb has before you grab the new xml (themoviedb also). Maybe a forced refresh anyway if its 60 days later?
• make sure xml sources gave good xml before you save
• make sure xml is good when you read it in case it needs a refresh
• don't delete the cache on a refresh until you have a new good transfer and use the old cache if you couldn't get one/the site was down.
• some log messages that it used a cached or fresh copy for those site changes that happen.

[Omertron]

Comment #1 originally posted by Omertron on 2010-10-27T14:24:53.000Z:

An afterthought, with horrible tv support for nfo managers it might be a great option to allow cache for type of media.. so you can not cache movies but cache tv shows..

[Omertron]

Comment #2 originally posted by Omertron on 2013-02-18T17:04:04.000Z:

This issue will be skipped.

Reasons:
1.) YAMJ has already a cache: the jukebox folder itself.

2.) Most sides use JavaScript or CSS to load images, tags and so on; so that must be cached also.

The best approach you could do is to install a proxy server like squid, which also has a functional and good cache mechanism (also for images). It's open source and available for nearly every platform and OS. Also you can set a limit when cache content will be deleted (size or days or whatever).

So you can just set the proxy settings in YAMJ and use the caching mechanism of SQUID.