Feature Comparison


Feature
MySQL Cluster
ScaleDB
Foreign Key Support
No
Yes
Index File Larger than Memory
No
Yes
Index type
B-tree
Multi-Table 1
Handles Large Transactions Well
No
Yes
Distributed Table Locks
No
Yes
Joins
Very slow
Very fast
Rollback Granularity
Entire transactions
Statements
Range Scans
Very slow
Very fast
Architecture
Shared-nothing
Shared-disk
Target Market
Telco
General Purpose
Supports Virtual Environments (e.g. VMWare, Cloud, etc.) No Yes

1. Unique multiple-table index provides dramatic improvement in join performance in addition to other benefits, including referential integrity and XML indexing.
Source for MySQL Cluster (NDB) information: http://download.oracle.com/docs/cd/E17952_01/refman-5.1-en/mysql-cluster-limitations.html and http://dev.mysql.com/doc/refman/5.1/en/mysql-cluster-ndb-innodb-engines.html

Comparing NoSQL, MySQL and ScaleDB

 

Feature
NoSQL2
MySQL (InnoDB)
ScaleDB
Transactions
No
Yes
Yes
Joins
No
Yes
Yes
Data Consistency
No (Eventual)
Yes
Yes
SQL Support
No
Yes
Yes
ACID Compliant
No
Yes
Yes
Supports MySQL Applications without Modification
No
Yes
Yes
Dynamic Elasticity (add/remove nodes on the fly)
Yes
No
Yes
High-Availability
Yes
No
Yes
Eliminates Data Partitioning
Yes
No
Yes
Master Only (no slaves, replication, slave promotion)
Yes
No
Yes

2. The NoSQL column represents an aggregate of NoSQL key-value stores. We recognize that some NoSQL applications are providing things like forced consistency for specific elements and other such enhancements, however, these result in tradeoffs in areas such as performance and elasticity. Some are even starting to provide limited SQL support, which is really a translation layer providing partial SQL emulation. We believe that NoSQL will do quite well in the OLAP space (e.g. Hadoop) where the ecosystem (applications, tools, trained DBAs) are much less critical, however, just like ODBMS, XML Databases, In-Memory Databases, etc. they will not make much of a dent in the OLTP market, due to a lack of compatibility with the ecosystem.