The MIND Server Architecture
The architecture and interaction between the various servers contained in
the MIND Project is relatively complicated. The complete functionality is
contained not within 1 or 2 applications, but rather 5 different
applications running concurrently and communicating with one another on a
machine. The CTN portion of the system contains the server portion that
the client can connect to as well as several other applications which
provide extra functionality to the client. This CTN server communicates
with and uses several miniSQL databases in order to process the client's
The basic layout is as follows:
The MIND Query Client sends a query request to the archive_server
application running on a remote machine. The archive_server application
performs various checks to determine whether or not a particular
Query Client is authorized to connect to the server. If it is authorized
the archive_server initiates a DICOM association so that data may be
passed back and forth. Once the association is established, the
archive_server parses the query information sent by the Query Client and
then sends the request on to the miniSQL server for processing.
Please keep in mind that this is but a brief overview, and the actual
interactions are much more detailed. They will be explained in further
detail by following the above links, or the links from the previous page.
Much of the details on the interaction between the servers and databases
of the system is taken from CTNs official documentation on the subject
which can be found at http://dicomctn.wustl.edu/DICOM/ctn-docs/users-guide.pdf.
This requires the Adobe
Provided in the MIND Project is a simple script that will start all of the
servers needed for a single archive system. This is packaged in the file
start_servers and requires that the
user have root level access to the machine it is being run on.