What is SonicBase?
SonicBase is an in-memory distributed relational database. SonicBase allows you to store and retrieve records in a
database with fast insert and retrieval speeds. You can use it for
long-term storage and more short-term storage if you just want to do some temporary analysis of data. You may store
anything from financial data to human resources data.
Open Source Database
A highly functional subset of features
is provided in the open source version of this
distributed relational database. The open source project is hosted on SourceForge
Additional features are available in the pro version for a modest fee ($250 for 8 cores per year).
SonicBase stores all data in memory. It uses a mix of on-heap and off-heap memory. Additionally, disk-based
snapshots and logs are used to achieve durability.
Records and indices are spread out across as many servers as you deem necessary. The data is transparently divided
and it is dynamically distributed as you add or delete records. You can scale SonicBase to billions
of records on commodity servers. See the documented benchmarks for more details. SonicBase Pro supports cross-shard
transactions and cross-shard joins. This is not a simple task for a distributed database.
SonicBase is an RDBMS that supports the SQL syntax. SQL support is accessed via the JDBC driver.
SonicBase is a distributed relational database that is cross-platform, supporting Linux, Mac, Cygwin and Windows.
With a 32 server, 4 billion record cluster with replication factor equals 1 we achieved 3.8 million batch record
inserts per second and
35 million range-based record reads per second. With this distributed database you can achieve linear scalability.
Being an in-memory database, Sonicbase can achieve impressive performance.
Automated Bulk Import
We have added the ability to automatically load your SonicBase database from another JDBC compliant database. This
you to quickly load SonicBase and do analytics on your data. Bulk import is available in the open source database.
See Automated Bulk Import Documentation
Streaming allows you to ingest data into the database and export it in real-time into external data stores. You
could, for example, hook into
a financial stream to provide up to the second quotes to your database. Analytics can then be performed on the data
in SonicBase. Or perhaps
you want to replicate your SonicBase database to an external data store. There are many possibilities. Streaming is
high-performance real-time data transfer.
You can connect to Kafka, Amazon Web Services Kinesis, Amazon Web Services Simple Queue Service and custom streams
providers. The streaming
consumer can load your SonicBase
database in real-time, while you perform analytics functions on your SonicBase database. See Streaming Consumers
You can hook Kafka, Amazon Web Services Kinesis, Amazon Web Services Simple Queue Service and custom streams
providers into SonicBase so that all inserts, updates and deletes will be published into your
streams. See Streaming Producers
This in-memory database now supports stored procedures. You can run your Java code inside the SonicBase servers. With your
you can build or filter results before returning to the client. These powerful procedures can be used to greatly
enhance performance by avoiding sending data over the wire. See Stored
Like what you see or not? Please send us feedback.