Jobs are created on the job canvas by dragging nodes from the palette and connecting them in a logical order. A node can have any number of incoming or outgoing connections, thus providing easy step-by-step backups and redundant output. Certain nodes can use named pipes instead of files to provide near-real time processing end-to-end in a job. This flexible architecture can easily be enhanced to create mediation solutions for any customer requirement.
- Disk C
These are program objects whose function is to communicate with each network element (switch, router, server, etc.) and collect data records. The standard Disk Collector and FTP Collector are capable of handling most collection needs. When conditions require a specific collection protocol, a number of custom collectors are available.
Stiletto® processing nodes manipulate the data stream. Both the File Processing node and the Block Processing node accept standard C++ library plugins. These nodes process incoming data either a file at a time or a block at a time respectively. Custom libraries can be written to handle such standard things as real time CDR file translation, duplicate CDR file detection, duplicate file detection, and CDR file content manipulation. The Pipe to File Conversion node provides the user a mechanism to take an incoming real time stream of blocked data and separate it into files based on custom criteria such as time and file size. Together these nodes provide an open ended framework to solve most mediation requirements.
- Disk D
At any point in a job, output can be distributed to local disk or to another computer. The Disk Distributor can write a file to a directory on a disk local to the Stiletto® mediation system or to a disk on another computer on the same network. The FTP Distributor is used to send processed files over a TCP/IP network to downstream systems that support the FTP client.