MongoDB vs CockroachDB performance Test

I want to benchmark latency performance of mongoDb with cockroachDB. We have to build a transaction management system which can support scalability, consistency and low latency. It seems MongoDB also provides these functionality at some level . Can anybody let me know any benchmark results for cockroachDB vs MongoDB in terms of latency . Just need to get an idea which will perform better and by what percentage .

Hello! Before evaluating both platforms based on performance (I couldn’t find a good benchmark after a cursory search, but can look around further if you still wanted this), have you considered whether your app would be better suited to use a NoSQL database (Mongo) vs a Distributed SQL Database (CRDB). They provide very different developer experiences (Mongo doesn’t have a SQL interface for example) and consistency guarantees. See: CockroachDB in Comparison | CockroachDB Docs

CockroachDB can be a good fit as we want scalability and consistency
But with all this we also want low latency.
Consider we are having two table which is like


UserTable might have data approx 50 GB
UserTransaction might have data approx in PB which will continue to grow .
Queries would be as follows

  1. Get balance of user through userid (Single row returned)
  2. Get transaction through transactionId (Single row returned)
  3. Get all transactions of user through userid (Multiple row returned approx 10,000)

Just want to have general idea how cockroachDB will perform in this usecase . Basically want traffic QPS vs latency number . for e.g : for single row return at 100 QPS traffic P99 latency is 50 ms . It is just the hypothetical number/example I am quoting here :slight_smile: .

We have some standardized performance benchmarking numbers using the TPCC workload. Read about those test results here: CockroachDB Performance | CockroachDB Docs