High-Speed Capture

Figure 1 describes how the Transaction Server fits into a high-speed processing environment.

Figure 1. The Transaction Server in a high-speed processing environment

Data Flow

This section describes the data flow through the processing environment for high-speed capture. The network connectivity that exists for the current check processing environment has been highlighted in italics and details of this network connectivity have not been specified.

CPCS communicates with the Image Server at the start of the prime-capture process. This is a bi-directional data flow, and the data volume does not exceed 1K in either direction. The number of messages exchanged is between five and eight in each direction, depending on the type of startup (new entry or restart). This data transfer is network-speed critical.
The Image Server then communicates with the camera. This is a bi-directional negotiation process where the volume of data in either direction does not exceed 1K. The number of messages exchanged is one in each direction and will not exceed two. This data transfer is network-speed critical.
CPCS communicates with the sorter, and the capture process begins.
As documents are read by the sorter, the MICR data is sent to CPCS.
As documents are scanned, the image camera sends an image record, which includes the MICR data, to the Image Server. The size of this message is the image size + 1716 bytes of control data. The Image Server sends a 100-byte image acknowledgment record back to the image camera for each image it receives. This is bi-directional message exchange and network speed critical.
Steps 4 and 5 are repeated until the capture is complete.
The images and MICR data received at the Image Server is stored to image file sets. As the number of items in the image file set reaches a customer-configured maximum value, the image file set is closed and a new one is opened.
The Archive Server is notified as each image file set becomes available. The available image file sets are then read, and the image index and MICR data for each document in the image file set are sent across the network to the Transaction Server. Depending on your environment, this information may take two network hops -- the first hop from the Image Server to the Archive Server, and the second from the Archive Server to the Transaction Server.
What is important to note here is that only the image index and MICR data are being sent from the Archive Server, not the actual images. The total data size from the Archive Server to the Transaction Server does not exceed 1K. This is a bi-directional process, during which the Transaction Server sends acknowledgment data back to the Archive Server (approximately 100 bytes) as it receives each document.

The capture process is not dependent on the data transfer from the Image Server to the Archive Server. The only critical network considerations are from a business perspective -- how quickly must the data get into the Transaction Server.
CPCS communicates with the Image Server at the end of the prime-capture process. This is a bi-directional data flow, and the data volume does not exceed 1K for each message in either direction. The number of messages exchanged is five in each direction. This data transfer is network-speed critical.
Finally, at the end of the capture process, the Transaction Server sends a 1K data record to CPCS and receives a 1K acknowledgment back. This transfer is not critical to the capture process.

To summarize, at the end of all the above processing, there are images on the Image Server, and MICR and image index information on the Transaction Server.