I’m building an app and I’m thinking of using Elasticsearch. Someone recommended me to look into rockset. I have never used either solution but I’ve attended a few tech talks about ES before. What’s the fundamental difference between you and ES?
Hey Gerry! First, welcome to the Rockset community! My name is Rafael and I work in Product Management here. Let me try to give you all the information you need and if you are interested you can take Rockset for a spin with $300 free trial credits. Our team would also be happy to give you a demo (just signup at rockset.com) if you are interested.
I’m sure you did some google searches and found https://rockset.com/elasticsearch-vs-rockset-guide.pdf or maybe our blog series. While we made lots of improvements since that whitepaper came out, and so did Elastic to be fair, the fundamental differences are more or less the same:
- If you plan to manage your own Elastic cluster (using the open-source version), there’s obvious overhead here, but even the managed cloud version is still something you’ll need to monitor closely, plan and select the right hardware as you scale things. Rockset is serverless and has decoupled compute/storage, allowing you to scale at a click of a button, making operations a lot easier. (more details here https://rockset.com/blog/elasticsearch-rockset-real-time-analytics-managing-clusters-going-serverless/)
- Elastic supports only one type of indexing commonly referred to as search index, built for log analytics and text search where Rockset indexes your data by row, column, and inverted (key-value aka search). This gives you a lot more flexibility in running sub-second queries especially when you need to do aggregations
- Two really important points when it comes to querying your data that a lot of our customers care about are: JOINs and our Smart Schema. With Elastic, there are no JOINs and you typically need to run an ETL for your data, where Rockset can infer data types and adjust your schema as new data comes in (more details here https://rockset.com/blog/elasticsearch-rockset-real-time-analytics-query-flexibility/ )
- Last, if you care about how fast new data gets ingested, indexed, and available for your queries, Rockset offers better latencies in that respect, making it a much better fit for analytics on real-time data for applications, where with Elastic these latencies can be much higher. Granted that for Elastic’s typical use cases it might be less critical. (see more details here - https://rockset.com/blog/elasticsearch-rockset-real-time-analytics-ingestion-indexing/)
Feel free to follow up with me here or use any of the links above to see a demo or try us out. Happy hacking!