Note: Descriptions are shown in the official language in which they were submitted.
CA 02272425 1999-OS-19
1
Processor-Controlled System and Method
of Operating a Processor-Controlled System
This invention relates to a method of operating a
processor-controlled system as set forth in the
preamble of claim 1, particularly a network management
system, and to a processor-controlled system as set
forth in the preamble of claim 5.
For many applications, systems are controlled by one
or more processors executing a control program, such
as an operating system. For the control program, an
architecture of hierarchically arranged program layers
is frequently chosen, with each program layer
providing different services, and at least part of the
services of a higher layer building on services of a
lower layer. The layered architecture is used
particularly for network management systems.
In an article by M. P. Bosse et al, "Management von
SDH-Netzelementen: eine Anwendung der
Informationsmodellierung", Elektrisches
Nachrichtenwesen, 4th Quarter 1993, pp. 329-338, the
structure of a control program for a network
management system of an SDH network (SDH = synchronous
digital hierarchy) is described. The control program
consists of different, hierarchically arranged program
CA 02272425 1999-OS-19
2
layers. These program layers comprise a network layer
and an element layer. The article also mentions that
the network management system has database
capabilities. Communication between the program layers
takes place via a defined interface referred to as a
"Q3 interface".
According to an article by S. Colombo, "Technologie
zo
der SDH-Netzelemente", Elektrisches Nachrichtenwesen,
4th Quarter 1993, pp. 322-328, the element layer is
also a control program with a layered architecture.
The network element functions reside in a first
program layer in the form of managed objects. This
program layer builds on functions which are provided
by the program layer referred to as "virtual hardware
module". Located below the virtual hardware module is
the on-board controller software, which accesses the
20 SDH hardware. The first program layer incorporates a
permanent ("persistent") storage for storing
configuration parameters (managed objects) in a
database referred to as a "persistent database".
With the layered architecture described, communication
between adjacent program layers is necessary: A higher
program layer must forward requests to a lower program
layer, and the lower program layer must return a
confirmation of the execution of the request and, if
30 necessary, a result. A higher program layer generally
waits for the confirmation, and the lower program
layer does not send the confirmation until the request
has been successfully executed.
A problem associated with this approach is that with a
CA 02272425 1999-OS-19
3
layered program architecture, the entire system may be
blocked when a higher program layer is waiting for a
confirmation from a lower program layer. when the lower
program layer is busy and does not get around to
executing the request received from the higher program
layer. The throughput through such a multilayer
control program is thus limited by the speed of the
slowest layer. Separating ("decoupling") the program
layers by means of queues does not result in a basic
improvement, since queues may overflow, for example in
the event of a failure of a program layer due to an
error.
It is an object of the invention to provide a method
of operating a processor-controlled system,
particularly a network management system, as well as a
processor-controlled system wherein blocking cannot
occur as a result of a busy condition or failure of a
program layer.
This object is attained with respect to the method by
the features of claim 1 and with respect to the system
by the features of claim 5. Further advantageous
aspects of the invention are defined in the dependent
claims.
One advantage of the invention is that in the event of
an error, the availability, robustness, and
survivability of a system according to the invention
is improved.
Another advantage is that the efficiency of a system
according to the invention is improved, since several
CA 02272425 1999-OS-19
4
requests can be combined and processed together in one
program layer. In addition, throughput through the
individual program layers is not limited by the
processing speed of the slowest program layer.
A further advantage is that the system according to
the invention remains operable at least externally
even in the event of a total failure of a middle or
lower program layer. A special case of a total failure
exists if part of the system is shut down or not
connected. The invention thus permits a transparent
off-line configuration of systems according to the
invention.
A preferred application of the invention is in the
area of network management, namely as a network
management system of a telecommunications network or
as a controller of a network element of a
telecommunications network.
The invention will become more apparent from the
following description of an embodiment when taken in
conjunction with the accompanying drawings, in which:
Fig. 1 shows schematically the structure of a
system in accordance with the invention;
Fig. 2 shows the data flow through a program
layer of a system in accordance with the
invention; and
Fig. 3 is a flowchart of the method in accordance
with the invention.
To separate ("decouple") individual layers of a
' CA 02272425 1999-OS-19
control program with respect to the time sequence of
the program run, according to the invention an
asynchronous approach is chosen. A basic idea of the
invention is to first process a request locally in a
higher program layer and send back a confirmation, and
subsequently forward the request to a lower program
layer for further processing. This imparts to the
confirmation the character of a promise to take care
that the request will be executed. The confirmation
thus no longer states that the request was
successfully processed, but acknowledges the receipt
of the request and promises to process it. By these
measures, effective, complete decoupling of the
program layers is achieved.
According to the invention, a user, in response to a
request, receives a confirmation in the form of an
"OK" before the request was actually executed.
Particularly advantageously, the request is logically
and semantically checked in a current program layer.
This ensures that the request is executable at least
in principle, and the confirmation then represents an
assurance that the request will be executed.
In order that local.processing of the request can be
performed, each of the program layers has a memory
associated with it in which the current configuration
parameters of the respective layer are stored and
saved. "Configuration parameters" as used herein also
means event and status data of the respective program
layer. The memories of the program layers are
logically separate memories which, however, may be
physically implemented in the same memory device, such
CA 02272425 1999-OS-19
6
as a RAM, an EEPROM, a hard disk, or another data
carrier. Preferably, the memory is structured as a
database in which the configuration parameters are
stored. The control program may also be complex
control software consisting of a number of program
modules which are executed as a distributed
application on different processors.
In the embodiment shown in Fig. 1, the system
according to the invention, SYS, consists of four
program layers LAYER1 to LAYER4. Each of the program
layers has a database DB1 to DB4 associated with it.
The databases contain configuration parameters of the
associated program layer. To obtain access to the
system SYS, a request REQ is sent to the highest
program layer LAYER4. In this program layer LAYER4,
local processing of the request REQ is performed by
updating the configuration parameters affected by the
request in the database DB4 in accordance with the
request. When this has been done, a confirmation CONF
is sent in response to the request REQ.
If the request REQ needs to be processed in the next
lower program layer LAYER3, it will be forwarded to
this layer. There, the same steps are taken: local
processing of the request by updating the affected
configuration parameters in the database DB3, and
subsequently sending a confirmation.
The forwarding to the respective next lower program
layer continues until the lowest program layer
necessary to execute the request is reached.
CA 02272425 1999-OS-19
7
This will now be illustrated by the example of a
network management system. The program layer LAYER4 is
an application layer in a central network management
facility. The next lower program layer LAYER3 is the
so-called Management Information Base (MIB, as
standardized in ITU-T X.720) together with a software
"framework" in a controller of a network element, for
example of a digital crossconnect. The next lower
program layer LAYER2 is a virtual hardware module VF~I
which performs the conversion between the MIB and the
hardware. The lowest program layer LAYER1 is the
firmware, which is held on the individual boards of
the crossconnect and controls the so-called on-board
controllers (OBC).
A request from a user to change a particular
connection parameter of an existing connection must be
processed by all four program layers. By contrast, a
request to read a connection parameter of an existing
connection can already be fulfilled by the first or
second program layer, depending on the type of the
parameter.
A special advantage of the invention is that several
requests can be combined and processed together in one
program layer. If, for example, the highest program
layer receives several requests which all want to
change, directly or indirectly, a particular
parameter, then the particular parameter will actually
be set to a new value in the highest program layer
several times in succession. This, however, takes
place only locally. When being forwarded, the several
requests can be combined into a single one, so that in
CA 02272425 1999-OS-19
8
each of the lower layers, only a single change needs
to be made to the respective parameter. This approach
is particularly advantageous in case of an off-line
configuration of the system according to the
invention, i.e., if the lower program layers are not
available. In that case, all changes will be made in
the local database, and after system start-up, the
changes necessary as a~result of the off-line
configuration will be forwarded in a single request to
the lower program layers.
Event reports can initiate consequent actions at the
next lower layer. This results in a further advantage
of the invention, namely that a request to perform the
consequent action can be sent to the next lower layer
even if the next higher layer, to which this event
report is to be forwarded, cannot currently process or
accept these event reports. Several event reports can
be combined if the higher program layer to which they
are to be forwarded cannot accept any reports for a
prolonged period of time. In this manner, only the
respective current status is forwarded. Events are
thus processed locally and, if possible, are combined
when being forwarded (adaptive event suppression).
Fig. 2 shows schematically the data flow in a program
layer. The program layer LAYER N receives requests REQ
from the next higher program layer, processes the
request locally, and sends as a response a
confirmation CONF. The request REQ is then forwarded
to the next lower program layer. In the reverse
direction, the program layer LAYER N receives event
reports EVE from the next lower program layer. Certain
CA 02272425 1999-OS-19
9
events may require consequent actions, so that
processing of the event report is performed in the
program layer LAYER N by sending to the next lower
program layer a request REQ to perform the consequent
action. In response to the event reports, the internal
database is updated if necessary. The event report EVE
is then forwarded to the next higher program layer.
Fig. 3 shows a flowchart of the method according to
the invention. The method comprises the following
steps:
Step S1: The system is accessed by sending a request
to the highest program layer. The highest
program layer thus becomes the current
program layer.
Step S2: The request is executed in the current
program layer.
Step S3: The configuration data in the memory,
preferably in the database, are changed in
accordance with the request.
Step S4: As a response, a confirmation is sent.
Step S5: If the lowest program layer needed to
execute the request has not yet been reached,
the request will be forwarded to the next
lower program layer for further processing.
The next lower program layer thus becomes
the current program layer. Then, steps S2 to
S5 are repeated.
"Current program layer" as used herein means the
program layer~which is currently processing the
request locally.
