search

Tyler-pierce / giza_sphinxsearch

Sphinx Fulltext Search Client for Elixir Projects (Phoenix Example Included)
19 stars 5 forks source link

Project Status #2

Closed firrae closed 3 years ago

firrae commented 4 years ago

Hi there,

This seems to be right up the alley of what I need. I had been looking for a package similar but I'm curious on this projects status? If its dead do you have an alternative i should look at that maybe you've moved to?

Let me know when you can, thanks.

Tyler-pierce commented 4 years ago

Hello In Elixir/Erlang there is no alternative that has been updated recently that I know of. I still use this project as the interface to sphinx and manticore search haven't changed much. You may want some niceties added but you won't reach a dead-end as you can always fall-back to sending a raw query and this lib will take care of some of the grunt work at least.

I'm still monitoring for any issues and will respond at the least to change requests. If you want to tell me more about your needs happy to let you know if this lib serves them.

firrae commented 4 years ago

Awesome. I'll try it out. I'm evaluating Manticore/Sphinx vs. Elastic Cloud for a small app I'm building so I'll let you know if run into anything along the way and work though it as possible.

I'm building a pretty simple app where I just need the ability, at the start at least, to do efficient fulltext search across a number of fields in a single table. For my "production" I'd like to be able to search across a couple different tables, but I'm willing to start walking before running even if that means scoping what the search entails via a UI element.

My biggest fear with ESC is cost, whereas it seems I could feasibly run Manticore/Sphinx on a $5 DO node until I saw performance issues.

Tyler-pierce commented 4 years ago

Alright good luck and feel free to ask if you have any uncertainty with your tests. You are correct in that it's strength is running smaller/lighter. I've had no troubles myself on EC2 free instance running small indices.