Logistics problem


I have search algorithm to search unindexed content for huge amount of results. The problem is, if I wanted to do search in real time, the client would probably timeout as such search takes up to 10 minutes.

So I am wondering, what logistics to go for (easiest and most reliable).
Do I just schedule search and let server do it in background and notify user via email when complete? Or do I get it running asynced and let user wait for results?

Also, where do I store results? Should I do it in database (array with up to 10,000 elements) or in text file?

Help me, I am quite fresh to web stuff.

nah, back to step 1. It wont work. I need to re-thing this thing.

Sure, that’s why google takes 10 minutes to search the entire web for me :slight_smile:

Search and index them while your user isn’t looking, then provide (almost) real time results in milliseconds.

You should probably just index the content so your searches can complete quickly.

its a no-go, this is search for external content which I cannot manage. Its basically YouTube Peoples search, and I cannot do it otherwise than browsign keyword related video pages. And this takes time.