Multiplex communications – Fault recovery – Bypass an inoperative station
Reexamination Certificate
2000-09-11
2002-08-06
Ngo, Ricky (Department: 2664)
Multiplex communications
Fault recovery
Bypass an inoperative station
C370S245000, C370S248000, C714S717000
Reexamination Certificate
active
06430151
ABSTRACT:
FIELD OF AND BACKGROUND OF THE INVENTION
The invention relates to a local network having redundancy properties. More specifically, the invention relates to an Ethernet network having a redundancy manager.
Ethernet networks and their properties are known from German Laid Open Publication DE 19 513 316 A1, German Laid Open Publication DE 19 513 315 A1, European Laid Open Publication EP 688 121 A1, U.S. Pat. No. 5,469,503, PCT Publication WO 95 15 641 A1 and U.S. Pat. No. 4,800,559, for example. Generally, redundant communication in automation systems has previously been implemented by a dual structure of the entire automation systems. This dual structure includes subscribers and a network infrastructure (dual bus system).
European Patent EP 0 403 763 B1, the disclosure of which is incorporated into the present application by reference, teaches a linear Ethernet network, in which the two line ends of the network are connected to a redundancy manager. If there are no errors in the network, the redundancy manager separates the two line ends from each other. If errors do occur in the network, then the redundancy manager connects the two line ends together. The redundancy manager checks whether a transmitted data signal appears simultaneously at the two line ends within a predetermined period of time. Thus, the redundancy manager checks whether the linear network operates in an error-free manner.
However, in the error-free case, this testing criterion can only be applied in a linear network, in which transmitted data telegrams actually do appear at both line ends. This condition is fulfilled, for example, in a linear network that is structured with layer-1-components, i.e., with components that only carry out a regeneration or amplification of the telegrams but that do not carry out an address evaluation and a telegram routing. On the other hand, in a network with layer-2-components, e.g. bridges, which carry out telegram routing, i.e., analyze the target address and the source address of the telegrams and forward the telegrams according to their addresses to connected segments, a transmitted data signal does not necessarily simultaneously appear at both line ends, even in tile error-free case. Thus, this testing criterion is not always applicable in a linear network having layer-2-components.
European Patent Application EP 0 052 390 A1, the disclosure of which is incorporated into the present application by reference, teaches a network with redundant transmission lines, in which one of the transmission subscribers transmits, in a fixed time clock, test messages to the other subscribers. These test messages are transmitted in the form of test packets that are evaluated by a receiving circuit. If the test packets are error-free, a downstream evaluation logic generates a switching signal for the reception of signals from another bus line. If, however, errors did occur on a bus line, then the downstream evaluation logic generates a report signal to report this particular bus line.
Another possibility of media redundancy is provided by layer-2-components (bridge/switch) using the standardized spanning tree protocols (IEEE 802.1D). Since this protocol can handle network structures, which are interconnected in my desired way, it is relatively complex. Using the standard parameters in the spanning tree protocol, it takes roughly 30 to 60 seconds after occurrence or elimination of an error in the network until the network returns to a stable status, depending on the complexity of the network structure and the number of layer-2-components. By optimizing individual parameters, this time can be reduced. However, the minimum achievable reaction time is never shorter than some 5 seconds. This time interval, however, is unacceptable for automation systems, because, during this period of time, no productive communication can take place between the automation systems. In particular, the connected subscribers would dismantle logical connections. The process at hand would progress in uncoordinated manner. Alternatively, an emergency cutoff of the automation system would have to be carried out.
OBJECTS OF THE INVENTION
It is one object of the present invention to provide a fast media redundancy in a local network, particularly in an Ethernet network, and to provide a suitable redundancy manager. Fast media redundancy means that, after occurrence or elimination of an error, the network reconfigures itself into a functioning structure in less than a second. This is a critical time period for automation systems.
SUMMARY OF THE INVENTION
This and other objects of the invention are achieved by providing a local network having redundancy properties and by providing an associated redundancy manager and a method for detecting and eliminating errors in the network. The network includes a line, a redundancy manager, and switching modules, which are configured as layer-2-components. The line forms a line-shaped topology of the network and has two line ends. These two line ends are connected to the redundancy manager.
If there are no errors in the network, the redundancy manager separates the two line ends from each other. If there are errors in the network, then the redundancy manager connects the two line ends with each other.
The redundancy manager feeds, in first predetermined time intervals, at least one first test telegram into the two line ends. Thereafter, the redundancy manager derives a first command to separate the two line ends, if the first test telegram is received at a respective other one of the two line ends within a second predetermined time interval. If the first test telegram is not received within the second predetermined time interval, then the redundancy manager derives a second command to connect the two line ends.
The network topology includes a line having layer-2-components. The redundancy manager is preferably designed as a layer-2-component, which has special operating software to control the network structure and which is coupled to the two line ends of the network. As noted above, in the error-free case, the redundancy manager separates the two line ends from each other. If errors do occur, for example when a line is interrupted or a layer-2-component fails to operate, the redundancy manager connects the two line ends together. In this way, a functioning line is provided again. For a rapid, reliable error-detection in the network or for error-elimination, two mechanisms are used. These mechanisms are particularly advantageous if they are used in combination.
Error Detection Mechanism 1:
Each layer-2-component within the network detects failures of the data line or failures of an adjacent layer-2-component as well as the elimination of these errors with the mechanisms standardized in IEEE 802.3. After an error is detected, the layer-2-component sends a special “error occurred” telegram to the redundancy manager, in order to report the detected error. After an error is eliminated, the layer-2-component sends a special “error eliminated” telegram to the redundancy manager, in order to signal the change in the network structure.
Error detection mechanism 2:
In predetermined time intervals T
1
, the redundancy manager feeds test telegrams into the two line ends. If the test telegrams are received at the respective other line ends, the redundancy manager generates a command to separate the two line ends. If the test telegrams are not received at the respective other line ends within a predetermined interval T
2
=n·T
1
, i.e., if the test telegrams on the line are lost, the redundancy manager recognizes an error in the network. If the T
1
and T
2
parameters are appropriately selected, the error detection takes place in a time interval that is considerably shorter than one second.
After connecting the line ends, i.e. after the occurrence of an error, or after separating the line ends, i.e., after elimination of an error, the redundancy manager sends a special “network topology change” telegram to the other layer-2-components to inform them of the change in the network topology. Aft
Glas Karl
Lohmeyer Joachim
Reuschen Rolf
Schmid Bernhard
Ngo Ricky
Siemens Aktiengesellschaft
Sughrue & Mion, PLLC
LandOfFree
Local network with redundancy properties having a redundancy... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Local network with redundancy properties having a redundancy..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Local network with redundancy properties having a redundancy... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2877926