search

pmazurek / aws-fuzzy-finder

:zap: SSH into instances using fuzzy search.
MIT License
129 stars 27 forks source link

First try with simple redis as caching layer. #9

Closed nvtkaszpir closed 8 years ago

nvtkaszpir commented 8 years ago

Actually my first try with redis, but it works.

pmazurek commented 8 years ago

I am afraid adding redis as a dependency is a little too much. I think it would be ok to just store the data somewhere in a file.

pmazurek commented 8 years ago

I mean this would require someone to actually spin a whole database in order to use just one functionality. It seems like a heavy overkill. Nonetheless I'm tempted to merge this PR, because its strong work, and then just switch redis to be something file-based.

pmazurek commented 8 years ago

Ok I'm merging it, once I'm done modifying it, I'll release a new version. Thank you so much for this contribution @nvtkaszpir , amazing work!

nvtkaszpir commented 8 years ago

Hm, you're right, redis may be an overkill and plain files should be enough. Now when you mentioned it I think this should go to separate branch or as separate module for caching.

Why introducing redis: With certain amount of people we sometimes start to hit AWS API calls limits. So I was thinking about using shared database among people at work to store data, and this is a first attempt to implement such storage, there are no critical data. In general we would set up separate redis server, a worker to query AWS regions once a minute (should be enough), so people would be able to get require data faster. Yes, this may be an overkill for single user.