A time-series database for time-series data
November 8, 2007
Recently, Logical Information Machines (LIM) released a new 64-bit version of Historis, a database that accepts only time-series information and focuses on optimizing its storage, manipulation, and retrieval algorithms specifically for time-series data. This new version is virtually unlimited in the amount of disk space that can be used, offers faster speeds and improved analysis capabilities.
Time-series data is any numerical value that can be attached to a date and time. Most organizations use relational databases for their time-series data, which means they have to maintain a separate set of related files. Because relational databases handle text, have no standard organization or table structure, they often become jammed when faced with a lot of data. Historis handles only time-series data, with an insertion rate of more than 250,000 records per second and an extraction rate of more than 220,000 records per second. Unlike most relational databases, Historis maintains these rates regardless of the volume of data being managed.
With no limit in the amount of disk space used, the size of the database can grow as large as the disk farm can grow. With an estimated exabyte capacity, Historis users arent going to run out of room any time soon. In addition, Historis utilizes Natural Compression, thus reducing the amount of disk space required. Steve Johnston, Director of Server Software at Logical Information Machines, states, What takes a terabyte in a relational database may only take a couple hundred gigabytes in Historis.
Historis comes with its own user-friendly query language called MQL. An MQL query looks very much like English, and can be read and understood by anyone, without having to learn a programming language.
Several Analytic Tools are provided in Historis. The first, XMIM is a tool that helps the user build queries. MIMIC is a simplified XMIM that hides the queries, keeping things simple for the users. In addition, an Excel add-in lets users extract data into Excel.
Historis is a client/server architecture that meshes well with IT environments. It is set up inside a LAN allowing the user to run various client applications off Windows or Mac desktops or UNIX workstations.
A full array of APIs, including C++, VB.Net, C#, and Java is provided so that users are also free to write their own application clients. Historis can be combined with all major relational databases allowing data to flow between Historis and relational database solutions.
Earlier this week Logical Information Machines (LIM) and Team and Concepts (TnC) announced that they have joined forces, offering users of the recently released Historis 64-bit database the ability to exchange data using the online spreadsheet EditGrid.
For additional information, please visit the Logical Information Machines website.