DEV Community

Cover image for Elasticsearch vs Meilisearch
adithyasrinivasan
adithyasrinivasan

Posted on • Edited on

Elasticsearch vs Meilisearch

During the process of building arbeitnow there was a need for providing a near instant search experience for Job Seekers.

Criteria:

  • Easy to integrate with PHP
  • Runs reasonably well on VPS systems
  • Good community support (GitHub or Stackoverflow)
  • Flexibility

I had a bit of experience on Elasticsearch with a small project, which I learnt with the help of Udemy tutorials which were great. It should have been the obvious choice, except for the fact that it does require a steep learning curve and setup to get it going smoothly.

So I decided to give Meilisearch a try. It's quick, small footprint, integrates well using their SDK and provides great tolerance for typo's. But unfortunately, it lacks a lot when it comes to customising Search filters.

Things missing from Meilisearch:

  • Setting up Synonyms from a file
  • Multiple WHERE clauses, thus limiting search to simple queries and restricts us to mostly single columns WHERE conditions
  • Multi Index searching
  • Sorting Results

Elasticsearch supports all of the above (their recent licensing change is another thing!) so I decided to go with it.

Meilisearch's public roadmap is indeed promising, so I will surely be coming back to it in the near future, it is a great tool for smaller projects and hopefully it gets the love it deserves!

Top comments (2)

Collapse
 
atapatel profile image
Ajay Patel

I agree. Melisearch can be a good alternative to Sphinix. It has to achive more to reach at current level of Elastic search.

Collapse
 
adithyasrinivasan profile image
adithyasrinivasan

Yep, I am keenly watching how it progresses. It cannot compete with the beasts yet but can do well if your scope is limited.