Home > Why QuasarDB > Comparison

Unique feature set

Quasardb is high-performance key-value store, what sets it apart from the rest?

  • Scale up and scale out
  • Small footprint
  • Hot plug’n’play cluster growth
  • Accepts entry of any form and size
  • Batch requests
  • Concurrent distributed transactions
  • Concurrent distributed containers
  • Advanced operations (atomic compare and swap, get and remove, etc.)
  • And much more!

Performance

If you are looking for a key-value store, it is very likely you care a lot about performance. Rest assured that quasardb delivers top of the line performance while using very few resources.

Quasardb can deliver extremely high throughput on reasonable configurations, resulting in an excellent total cost of ownership (TCO).


Hardware configuration of the above benchmark:

  • In-memory analytics server: Cisco UCS C240M4 - 2 Intel E5-2697 v3 - 256GB – 2 x 40 GB VIC 1385
  • quasardb cluster: Cisco UCS C240M4 server - 2 Intel E5-2697 v3 - 256GB - 2 x 10GB VIC - 24 1TB DD running Linux
  • Switch: Cisco Nexus 9372 PQ (2 40GB connections to the simulation server – 8 10GB connections to two fabric interconnects)
  • Fabric interconnect: 2 x Cisco UCS 6248 equipped with the UCS manager software

Wherever you want to go, we can take you there.

But don’t take our word for it, see it for yourself!

Comparison matrix

We’ve selected a subset of NoSQL engines, mostly key/value stores but also document stores to give some more perspective and done our best to fill that comparison grid. Feel free to contact us if you believe some information is invalid or outdated.

Data model
Quasardb
Key/value
Aerospike
Key/value
Cassandra
Key/column
Couchbase
Document and Key/value
Mongodb
Document
Redis
Key/value
Minimal configuration
Quasardb
64 MiB RAM, 1 core, block based storage device
Aerospike
4 GiB RAM, 4 cores CPU, SSD persistence
Cassandra
4 GiB RAM, 8 cores CPU, 2 independent disks
Couchbase
64 GiB RAM, 4 cores, block based storage device
Mongodb
Unspecified+
Redis
8 MiB, 1 core, block based storage device
Persistence
Quasardb
Synchronous
Aerospike
Asynchronous, SSD only
Cassandra
Asynchronous
Couchbase
Asynchronous
Mongodb
Asynchronous
Redis
Asynchronous
Implementation language
Quasardb
C++ 14
Aerospike
C
Cassandra
Java
Couchbase
C, C++ 03 and Erlang
Mongodb
C++ 03
Redis
C
Supported OS
Quasardb
FreeBSD, Linux, OS X and Windows
Aerospike
Linux, OS X and Windows
Cassandra
Linux and OS X
Couchbase
Linux, OS X and Windows
Mongodb
Linux, OS X, Solaris and Windows
Redis
UNIX/Linux
Scale up
Quasardb
Yes
Aerospike
Yes
Cassandra
Yes
Couchbase
Yes
Mongodb
Limited
Redis
No
Maximum entry size
Quasardb
No internal limit
Aerospike
128 kiB (strings) +
Cassandra
10 MiB++
Couchbase
20 MiB
Mongodb
16 MiB
Redis
2 GiB
Clustering
Quasardb
Peer-to-peer
Aerospike
consistent hashing
Cassandra
consistent hashing
Couchbase
consistent hashing
Mongodb
Master/slave
Redis
consistent hashing
Replication
Quasardb
Yes
Aerospike
Yes
Cassandra
Yes
Couchbase
Yes
Mongodb
Yes
Redis
No
Atomic integers
Quasardb
64-bit signed
Aerospike
64-bit signed
Cassandra
32-bit signed
Couchbase
N/A+
Mongodb
N/A
Redis
64-bit signed
Queues
Quasardb
Distributed
Aerospike
Lists
Cassandra
Lists, not distributed
Couchbase
Arrays, not distributed
Mongodb
N/A
Redis
Not distributed
Sets
Quasardb
Distributed
Aerospike
No
Cassandra
Yes
Couchbase
No
Mongodb
N/A
Redis
Not distributed
Multi-entry transactions
Quasardb
Yes (MVCC)
Aerospike
No
Cassandra
No
Couchbase
No
Mongodb
No
Redis
Limited
Schema
Quasardb
None
Aerospike
Namespace
Cassandra
None
Couchbase
Buckets/documents
Mongodb
Document-oriented
Redis
None
Automatic indexing
Quasardb
None
Aerospike
None
Cassandra
Column based
Couchbase
Document based
Mongodb
Document based
Redis
None
Secondary indexes
Quasardb
Unlimited tagging
Aerospike
Integers and strings only
Cassandra
Column based
Couchbase
Views
Mongodb
Multiple
Redis
None

+ documentation and testing didn’t provide a clear answer
++ recommended limit

Try now Contact