MamBase Research Ltd
Israel
PO Box 2527,
53 Shenkar st.,
Holon 58125, Israel
Phone: +972 53 303-3903
MamBase
High-performance universal database for modern IT solutions:
Support for relational, graph and document data models
High performance on any data model
Scalability, reliability
MamBase and Neo4j performance сomparison
Database query execution time (milliseconds)
The results have been verified by an independent system integrator
MamBase
Versatility
- Relational (SQL), graph (Graph-SQL) and documentary (JSON) data processing.
- Fast response on all data models.
- SQL, Graph-SQL and JSON queries to the same data.
- Mapping built into the core allows for data translation between the models.
Scalability, reliability
- High-performance MamBase cluster
- High-performance distributed queries on any data: distributed graph, relational JOIN queries
- High availability and fault tolerance
- Fault-tolerance server keeps the queries running when some of the cluster nodes fail
Architecture
MamBase supports a single repository for relational data, graphs and JSON documents. You can perform SQL, Graph-SQL and JSON queries on the same data.
Relational model
The MamBase architecture is designed to support the relational data model and SQL. Relational model represents data in tabular form.
SQL query
select
model, year, name
from
car join person on
car.owner = person.id
select
model, year, name
from
car join person on
car.owner = person.id
Graph model
Specialized structures are designed to represent graph data. When mapping graph data, the nodes are records in the relational tables. Relational links Foreign Key -> Primary Key can automatically be interpreted as the graph edges. The intermediate tables for many-to-many relationships are also visible as graph edges. Special indices significantly accelerate the graph query processing, as well as SQL JOIN operations.
The Graph-SQL language is developed for graph data processing. But since the graph can also be presented as relational data, SQL can be used for graph data queries.
Graph-SQL query
SELECT p1.name, p2.name as friendname, c1.model
MATCH (car c1)-[owner]->(person p1)-[friends]->(person p2)<-[owner]-(car c2)
WHERE c1.model = c2.model
SELECT p1.name, p2.name as friendname, c1.model
MATCH (car c1)-[owner]->(person p1)-[friends]->(person p2)<-[owner]-(car c2)
WHERE c1.model = c2.model
SQL query
SELECT p1.name, p2.name as friendname, c1.model
FROM car c1
join person p1 on c1.owner = p1.id
join friends f on f.fromid = p1.id
join person p2 on f.toid = p2.id
join car c2 on c2.owner = p2.id
WHERE c1.model = c2.model
SELECT p1.name, p2.name as friendname, c1.model
FROM car c1
join person p1 on c1.owner = p1.id
join friends f on f.fromid = p1.id
join person p2 on f.toid = p2.id
join car c2 on c2.owner = p2.id
WHERE c1.model = c2.model
JSON documents
JSON documents are automatically represented as records and relationships reflecting the nesting of objects or relationships between objects.
When you add this document, it is automatically presented in a relational form as a set of records in some tables, and also as part of a graph.
The same SQL and Graph-SQL queries can be used to process data received as JSON..
JSON document
{
"type":"car",
"id":1,
"model": "Chevrolet",
"owner": {
"type": "person",
"id": 1,
"name": "John",
"city": "London",
"friends": [ 2, 3 ]
}
}
{
"type":"car",
"id":1,
"model": "Chevrolet",
"owner": {
"type": "person",
"id": 1,
"name": "John",
"city": "London",
"friends": [ 2, 3 ]
}
}