The ANKHOR Enterprise Server is a modular server design and consists of 4 Server components. These can be used individually or combined to meet all the requirements of different data projects.
The ANKHOR server version 2.0 products are available in spring of 2015 for the operating systems Windows and Linux.
Typical server configurations can be found here .With the Web and Application Server (AHTTPS) several users can apply the applications modelled with ANKHOR FlowSheet Workstation from a browser or mobile device. Without any changes to the modelled applications, AHTTPS automatically generates HTML, so that the browser takes over the role of the Dashboard. In addition it can be deployed as a middleware server and operate as a pure web service with e.g. JSON. Its user management allows controlling access rights from the FlowSheet program.
FlowSheets can be uploaded to the AHTTPS server directly from a local development environment like ANKHOR Workstation without detouring through external data transfer protocols. Moreover, during their execution on the server they can be analyzed and controlled with the aid of the ANKHOR collaboration functionality.
Integrating FlowSheets with other web services is possible at any time by using them through the JSON service or an IFrame interface. Data export and data import directly from the web page can be implemented by appropriate input elements. In conjunction with the ANKHOR HTTP operators this results in a simple way to automatically process data from web services for presentation or visual analysis.
The ANKHOR Remote Execution Server (ARES) extends the ANKHOR server system with the ability to store data at a central location, share it with others and directly process it at this centralized location. ARES combines the ANKHOR execution engine with a persistence layer and management for users, access rights and transaction control and uses the features of the ANKHOR Compute Grid (AGRID). The major part of the data can remain in ARES, so that a powerful server (server park) with large main memory can be used by various users in common. Access to the data is done by libraries and is, for example, integrated into the TimeSeriesAnalysis package.
Thus ARES is an In-Memory No SQL Database which harmonically cooperates with the other ANKHOR components without any disruptions or obstructions caused by different (scripting) languages.
Usually queries to an ARES instance are processed with ANKHOR operators that are not executed on the querying machine but on ARES itself. This means that not the data but the algorithms are moved - a concept called data gravity.
A common application of an ARES system is implementing an analytics mirror server for an existing database. It is updated continuously or periodically and then provides the data to complex analytics functions without inducing extra load onto the operative system.
The ANKHOR Grid Compute Server (AGRID) can distribute FlowSheet computations onto several computers of a local network. For this, the Grid Compute Server is started on all computers intended to support the computation and together forming the Compute Grid, which can be used by the servers or ANKHOR instances outside the grid, like the Desktop Application Workstation.
The nodes (AGRID server) of an ANKHOR Compute Grid find each other through the name or address of one or several root nodes. Outage of one or several nodes does not affect the result since computations are moved around to other nodes dynamically. Thus the grid reacts in a flexible way to appearing or vanishing nodes and can also be distributed onto the workstations of a work group.
The ANKHOR Collaboration Server (ACOLLABS) extends the ANKHOR server system with the ability to execute FlowSheets on a central server but modify it from a local instance like ANKHOR FlowSheet Workstation. The FlowSheets of the collaboration server are stored in an associated ANKHOR Remote Execution Server (ARES), which also manages the access rights. Users with the appropriate rights can create new FlowSheets on the collaboration server using their local ANKHOR FlowSheet instance, or cooperate with other users on the same shared FlowSheet.
A FlowSheet can be executed as a background process (daemon) and in addition automatically be started along with the collaboration server startup.
As the collaboration server runs on a cloud server it facilitates executing large FlowSheets controlled by portable computers with low power. The proximity to an ARES server offers an additional step towards execution with the data paradigms.
