Key Features
EXASolution is ideally suited for real-time big data reporting, analysis and advanced analytics. Just like the Bigstep bare metal cloud, EXASolution management is highly automated, making it easy to use and maintain.
- Bare metal access to memory for minimum latency on in-memory operations.
- Fully scalable from 32 to 256 GB of RAM at the click of a button.
- All-SSD storage available for all instance types.
- 40 Gbps aggregated connectivity for high throughput.
The World's Fastest Analytic Database
on the World's Highest Performance Cloud
- EXASolution is a masterless, shared nothing, linearly scalable, distributed MPP RDBMS database.
- Using in-memory computing, EXASolution can access any value within nanoseconds.
- The vertical/ columnar storage makes most queries access only a few columns, ensuring lower latency than horizontal systems.
- Data is compressed in memory to optimize access to large datasets.
- EXASolution is fully unicode - a datatype specifies the number of symbols not the size in bytes.
- By being tuning-free, EXASolution automatically monitors both the data and the queries, optimizes both the creation and maintenance of indexes and caches and uses a strong cost-based query optimizer to enforce data locality as much as possible.
- EXASolution can run processing inside the database - EXAPowerlytics can execute arbitrary java code on a MapReduce-like scheme and connect to HDFS.
We deploy the latest version of EXASolution and integrate it to scale with the underlying infrastructure.
The cluster has two types of InstanceArrays: license server and nodes.
The license server is a management interface and does not do any processing.
The data nodes handle the actual processing and storage of data. For optimum performance, data nodes need enough memory to store the hot data required for any given query. The drives allocated to data nodes should be larger than the available RAM and over 120GB.
License Server
Instance Type: FMCI 8.32
Solid Store Drive size: 40GB
OS: Centos 7.x
Instance count: 1
Data Nodes
Instance Type: FMCI 20.128
Solid Store Drive size: 250GB
OS: Centos 6.x, Instance count: 3
LAN Connections: 2 - 1 x internal,
1 x external LAN for outside the cluster connection
EXASolution on Metal Cloud scales vertically and horizontally with the size of the underlying InstanceArrays
- When data nodes are added or upgraded, EXASolution is notified and handles the distribution of the data sets across compute instances and drives.
- For vertical scaling, instance configuration can be changed at any time with a simple restart. No additional delays are required for the data to be transferred out of each individual host.
EXASolution - Natively Compatible with Enterprise Data Warehouses
EXASolution is used extensively as a high performance and cost effective replacement to traditional RDBMSs and EDWs.
Transitioning from an existing deployment to EXASolution can be done seamlessly as EXASolution uses full SQL 2008 syntax and can work as a drop-in replacement to many SQL databases like MySQL, MS SQL, Oracle DB or IBM DB2. It has standard JDBC, ODBC, ADO.net, MDX interfaces and Client SDKs.
Integrated with Big Data on the Bare Metal Cloud
EXASolution on bare metal cloud can be integrated with several other applications in the bare metal cloud big data ecosystem. This results in powerful data analytics environments:
Hadoop on Bare Metal Cloud: CDH
Integration with Hadoop can be used for ETL, data sanitation and data set filtering, based on HDFS and MapReduce.
Datameer on Bare Metal Cloud
Datameer can be used as an analytics and visualization frontend to EXASolution, as it enables business analysts and data scientists to work with large data sets in a spreadsheet format.