Sunday, 4 June 2017

MongoDB replacement - SparkR/ Sparklyr?

Currently I have my R application (Shiny) pulling its data from Node.js, Express.js and MongoDB.

Now my data is getting larger and MongoDB only allows 16MB per document. So I have to split my data into different documents and then join them together for my R application.

It does not seem ideal to me. So I wonder Apache Spark can be a good solution or a MongoDB replacement for big data?

SparkR has been around for sometime while Sparklyr seems interesting and promising.

Is it plausible to replace MongoDB with Apache Spark? Can I move all my data into Spark?



via teelou

No comments:

Post a Comment