Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
2001-07-19
2002-08-06
Baderman, Scott T. (Department: 2184)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C714S043000, C709S251000, C370S222000
Reexamination Certificate
active
06430700
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to computer networks. In particular, the present invention relates to a system and method for providing a protection protocol for fault recovery for a two line bi-directional ring network.
2. Background of the Invention
The need for cooperation among various computers has motivated the growth of efficient architectures of networks. One topology of networks is a ring network, such as FDDI and SONET.
FIG. 1
shows an example of a two line bi-directional ring network. The ring network
100
is shown to include nodes
102
a
-
102
g
. Each node is typically a computer with embedded processors and at least one network connection. Each node
102
a
-
102
g
is shown to be bidirectionally coupled to two neighboring nodes
102
a
-
102
g
via an inner connection ring
110
a
-
110
g
and an outer connection ring
108
a
-
108
g
. For instance, node
102
a
is bidirectionally coupled to nodes
102
b
and
102
g
. The example of
FIG. 1
also shows a problem
104
in the connection between node
102
b
and node
102
c
. When a problem is detected (such as a bi-directional line cut), the connection between nodes
102
b
and
102
d
wraps back upon itself, as shown by wraps
106
a
and
106
b
. In this manner, the connection problem
104
can be avoided.
In a conventional SONET Line Switched Network, the nodes on each side of the problem
104
will typically exchange messages with each other over the long path. In this example, node
102
b
would send a message to node
102
c
via connection
110
g
, and vice versa, via connection
108
g
. In a conventional SONET Line Switched Ring Network, the nodes on each side of the problem
104
will typically exchange messages with each other over the long path. The long path is the path that passes the other nodes in the ring network to reach a neighbor of the node that originated the message. For example, a long path from node
102
b
to node
102
c
would pass through nodes
102
a
-
102
d
via outer ring
108
a
-
108
f.
When both the nodes flanking the problem
104
receive the other nodeis message, then these nodes will typically perform a wrap. For example, when nodes
102
b
and
102
c
receive each otheris message indicating the detection of the problem
104
, nodes
102
b
and
102
c
will perform wraps
106
a
and
106
b
. In order to send these messages, each node that sends a message must typically know the identity and location of the receiving node. If the identity and location of the receiving node is not accurate, then there will typically be a failure to wrap.
In a conventional SONET network, each message sent by a sending node to a receiving node typically needs the identification and location of the receiving node to arrive at the proper destination. Accordingly, manual configuration is typically needed in each node to store the identity and location of each other node in the ring network in order to provide for communication between the nodes in the network.
A problem can arise when a new node is coupled into the ring network. Each node then typically needs to have its topology map of the ring network reconfigured such that the identity and location of the new node in the ring network can be included. Additionally, the locations of at least some of the original nodes may also change and these changes should also be included in each nodeis internal map of the network. Until and unless these new locations and identities can be included in a sending nodeis internal map of the network, messages sent by a node indicating that a problem has been detected will typically not arrive at the proper destination node. When these message fail to arrive at the proper destination, the required ring wraps will also fail to occur.
Likewise, if two ring networks are merged into one, each node will then typically need to have its internal map of the ring network configured to include the identities and locations of each node included in the merged ring network. These reconfigurations typically require the time and effort of a programmer or network administrator. Again, unless and until these new locations and identities can be included in a sending nodeis internal map of the network, messages sent by that node will typically not arrive at the proper destination node. In summary, for the protection mechanism to operate, each node needs to know the current ring map (current ring topology). Accordingly, required ring wraps will fail to occur.
In summary, for the protection mechanism to operate, each node needs to know the current ring map (current ring topology). What is needed is a system and method for providing fault recovery for two line bi-directional ring network that minimizes the need to keep track of other nodes in the ring network. Preferably, the system would not require reconfiguration of an internal map of the network when a new node is added to, or existing nodes are removed from the network. The present invention addresses such a need.
SUMMARY OF THE INVENTION
The present invention provides a protection protocol for fault recovery, such as a ring wrap, for a network, such as a two line bi-directional ring network. An embodiment of the present invention works in conjunction with a ring topology network in which a node in the network can identify a problem with a connection between the node and a first neighbor. According to the embodiment, when the problem is identified, the node sends a message identifying the problem to a second neighbor which is located at least one node away from the problem. The second neighbor then forwards the message to a third neighbor, unless the second neighbor is dealing with a situation that is higher in a hierarchy of situations than the problem described in the message by the original node. In general, if the second neighboris situation has a higher priority than the situation described by the original node, then the message is ignored and not forwarded. If, however, the message sent by the original node describes a situation with a higher priority than the situation being dealt with by the second neighbor, then, in general, the second neighboris situation is ignored, at least for the moment, and the original nodeis message is forwarded to the next neighbor. In general, a higher priority request preempts a lower priority request within the ring. Exceptions are noted as rules of the protection protocol.
The present invention provides a protection protocol that simplifies the coordination required by the nodes in a ring network. The nodes do not need to maintain a topology map of the ring, identifying and locating each node on the ring, for effective protection. Additionally, independently operating ring networks can be merged and the protection protocol automatically appropriately removes a protection, such as a ring wrap, to allow the formation of a single ring. It also provides for multiple levels of protection priority so that protection for a high priority failure, such as a physical break in a connection, removes protection for a low priority failure, such as a signal degrade, on another link.
A method according to an embodiment of the present invention for fault recovery for a ring computer network, the ring network including a plurality of nodes, is presented. The method comprises detecting a situation by a first node, wherein the first node is one of the plurality of nodes; sending a first message via a short path to a second node, wherein the first node is adjacent to the second node; and initiating a fault recovery procedure when the second node receives the first message.
In another aspect of an embodiment of the present invention, a method for adding a new node to a ring computer network, the ring network including a plurality of nodes, is presented. The method comprises detecting a situation by a first node, wherein the first node is one of the plurality of nodes; sending a first message via a short path to a second node, wherein the first node is adjacent to the second node prior to an addition of the new node; init
Bates Tony
Chin Hon Wah
Daruwalla Feisal
Suwala George
Tsiang David
Baderman Scott T.
Cisco Technology Inc.
Van Pelt & Yi LLP
LandOfFree
System and method for fault recovery for a two line... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for fault recovery for a two line..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for fault recovery for a two line... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2948356