Specifications

1 Architecture
MCS stands for “monitoring and control system”. MCS is defined in [1]. The purpose of this doc-
ument is to describe the architecture of MCS (not discussed in [1]) and to provide the next lower
level of design detail.
Figure 1 shows the MCS architecture, as well as interfaces to adjacent subsystems. First, note that
here we use the term “MCS” to refer to both MCS (the upper shaded region) as well as the M CS
Data Recorder (“MCS-DR”; the lower shaded region). Excluding MCS-DR for the moment, note
that MCS consists of three computers and two switches. These are described below:
• The Scheduler is a computer whose primary function is to issue commands and receive status
from other LWA subsystems. The command and status communications are though the Com-
mand Hub, using the “MCS Common ICD” [2] augmented by the corresponding subsystem
ICDs. The term “scheduler” is a bit of a misnomer, since it does not actually do scheduling, but
rather keeps time, issues commands necessary to implement instructions from the Executive,
parses the stream of incoming response messages. The Scheduler manages the Command Hub.
The Scheduler handles tasks that are extremely time-sensitive and that must be coordinated
on timescales down to milliseconds.
• The Executive is the computer which exercises top-level control over MCS as well as the
station. It is responsible for interpreting observation requests and, from these, generating the
data which becomes the content of command messages issued by the Scheduler. This includes
numerically-intensive operations such as computation of FIR filter coefficients. It updates the
MCS and station MIBs accordingly. The Executive manages the Gateway. The Executive
manages tasks that are moderately time-sensitive and that must be coordinated on timescales
down to seconds.
• The Task Processor is a computer which exists primarily to host applications which are not
“time critical” and therefore can be “offloaded” to reduce the processing burden of the Exec-
utive. The Task Processor is the primary interface with users, managing command line and
GUI interactions. Prominent among these are observation scheduling, including the reduction
of scheduled observation data from “user friendly” form to parameterized instructions that can
be interpreted by the Executive. The Task Processor is also responsible for the scheduling and
interpretation of internal diagnostics (both automatic or user-directed), and manages MCS-
DR. In general, the Task Processor handles tasks that do not need to managed at resolutions
of seconds or less.
• The Gateway is a managed switch which, under the control of the Executive, regulates the flow
of network traffic between the Shelter LAN and the various computers of MCS and MCS-DR.
The Gateway is managed to prevent traffic from the Shelter LAN from interfering with MCS
operation, as well as to provide a layer of security between the shelter LAN (and external
network) and MCS-internal devices. The Gateway is also used to regulate the flow of data
between the MCS-DR computers and users who are retrieving data or otherwise utilizing
MCS-DR computers.
• The Command Hub is a managed switch which, under the control of the Scheduler, regulates
the flow of network traffic between the Scheduler and subsystems, utilizing the “MCS Common
ICD” interface [2]. The Command Hub is managed in order to guarantee sufficient allocation
of bandwidth between connected devices, and to avoid any connected devices from interfering
with MCS operation by “jabbering” or exhibiting other problematic behaviors beyond the
control of MCS.
MCS-DR consists of 5 identical computers as shown in Figure 1. Four of these computers receive
data streams from DP corresponding to the output of each of the four station beams. The fifth
computer receives the DP output data stream corresponding to TBW/TBN output. Each MCS-DR
2