8

I was in the process of evaluating nosql databases and finalised on two. RavenDB and MongoDb. Which one is better and why? Which will help in writing more performance oriented queries.

LordHits
  • 5,054
  • 3
  • 38
  • 51
Ajai rajan
  • 163
  • 1
  • 8
  • What specific cases have you already tested on both platforms on your hardware? If you edit your question to provide specific examples you've tried, this might be answerable. – WiredPrairie Feb 14 '13 at 12:02
  • Basically I will be using appharbor hosted instances...want to have information specific to the indexes,the number of times db hitting on executing query,complex searches,text search etc – Ajai rajan Feb 14 '13 at 12:11
  • You need to provide a lot more detail in your question to get any sort of reasonable and helpful answers. – WiredPrairie Feb 14 '13 at 13:18
  • I have been worrying about the same thing. I have been evaluating both mongo and raven but can't find any comparisons of these two platforms. – Luke101 Feb 14 '13 at 14:34
  • Check out my feedback on RavenDB after actively using RavenDB + .NET Core for a couple of years - https://alex-klaus.com/ravendb-pain-and-joy/ – Alex Klaus Jul 10 '19 at 01:41

1 Answers1

9

Raven's install base is smaller. When you consider the two, consider your growth path. Look carefully at how MongoDb and Raven differ on shard keys. Look carefully at support and community around the product - these will kick in once you have production critical issues. See if the indexing strategy and map-reduce constructs satisfy your use cases. Look carefully at the storage and latency of your largest collections and indexes. Decide whether the transaction model fits your needs.

If your application is simple enough, your abstraction over the DB layer should enable you to switch should you change your mind. But for a complex and robust application, I'd recommend a proof of concept before a full-prod commit.

Nuk Nuk San
  • 550
  • 3
  • 5
  • 8
    I cannot imagine that it is possible to abstract away over RavenDB and Mongo for reasonably sized applications. +1 – usr Feb 17 '13 at 20:38