Redis-search as a competing product?


(Justin Treher) #1

Have you guys (Team Elastic) been following the development of redis search? I've been watching a number of their presentations for redis enterprise to solve geographically distributed multi-master writes and some of those presentations have touted a large cost savings due to the performance difference between redis and elasticsearch for searching and aggregations purely from the reduced requirements of computing time in the cloud. For large data sets they are using flash storage instead of pure in memory. We aren't specifically interested in redis for search, but as a developer that uses ElasticSearch almost daily, it certainly caught my attention.

I'm not trying to start a flame war between products here, I just wanted to know if there was anything to borrow or steal here if there is any truth to the performance argument.

They are definitely missing a ton of features that ES has and the json based API is much more beginner friendly in my opinion. I feel like it's kind of like ES circa 2011 vs SOLR.

https://oss.redislabs.com/redisearch/


(Mark Walkom) #2

We aren't all guys :slight_smile:

Elasticsearch can potentially compete with a number of different products. Cost is only one factor of that and ultimately you need to be able to technically prove this claim, one way or another.

Simply trusting something in a presentation, or conversely via a forum post, is not really a recipe for success.


(Justin Treher) #3

@warkolm Of course you aren't all men. That colloquialism is just a bad habit I suppose.

I'm not worried about the cost; I do care about performance a great deal. There are a number of kitchen sink products that have search, but it's usually a second class citizen - I'm looking at you RDMSs. I was just bringing it up in case redis happens to be different. I'm sure you all have a team of people that regularly look at the competition and this post is probably redundant.


(Doug Turnbull) #4

It’s not just performance. I help people make search tech decisions and really the most important factor is mindshare. Can you hire people that work on it? Will the community be big enough to provide support, blog about best practices, put on conferences, write books on the technology? Are there company(s) behind the technology that provide products and services? Is the technology constantly being improved by the community for the use cases you care about? Will the technology be around in 5, 10, 20 years from now?

It’s hard to build that up and hard to overcome an incumbent like Lucene

That being said, it’s less hard to establish technology in a niche that Lucene-based search doesn’t do well yet.

Doug


(Justin Treher) #5

That's a good point. I'm sure redis looks like assembly compared to elasticsearch to most users.


(system) #6

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.