Multiplex communications – Fault recovery
Reexamination Certificate
2000-04-17
2004-06-15
Ho, Duc (Department: 2665)
Multiplex communications
Fault recovery
C370S244000, C370S450000
Reexamination Certificate
active
06751188
ABSTRACT:
FIELD OF THE INVENTION
The present invention pertains to communications and networking. More particularly, the invention relates to the usage of resources in networking devices.
BACKGROUND OF THE INVENTION
FIG. 1A
shows an example of a prior art wide-area networking system serviced by ATM (Asynchronous Transfer Mode). A wide-area network (WAN) link
120
interconnects a first network
100
with a second network
110
. Each network has a plurality of nodes which may each contain switching devices that regulate data traffic to one or more user terminals. Network
100
is shown having nodes
102
,
103
,
104
,
106
, and
108
, while network
110
is shown having nodes
112
,
114
,
116
, and
118
. A first user terminal
105
is connected to node
102
of network
100
and a second user terminal
115
is connected to node
118
of network
110
. In order for user terminal
105
and user terminal
115
to communicate with each other, a call must first be established between them. This call may be switched through a plurality of nodes. One possible route for sending data from user terminal
105
to user terminal
115
is for data to go from node
102
to
106
to
108
and then across the WAN link to node
112
and node
118
, finally reaching user terminal
115
.
Each node has a controller device and switch which facilitate the calls through the node. The controller has processing, memory, and other resources to interpret, forward, and process messages and initiate other messages as appropriate, while the switch ordinarily handles the physical routing of messages among nodes and user terminals.
In a redundant high-availability private network-to-network interface (“PNNI”) routing protocol implementation, a redundant pair of controllers is used in the case of controller failure. Thus, each node has two controllers, one primary and one secondary. In order to ensure the proper functioning of the secondary controller in case the primary controller fails, it is desirable to test the secondary controller by switching the primary controller to a standby state and the secondary controller to an active state. The switching of controllers is also desirable for performing maintenance on the primary controller.
A secondary controller follows the state of the primary controller with a slight delay—for example, a delay of one second. During a switchover, all activity must be transferred from the primary controller to the secondary controller. However, because of the delay between the controllers, synchronization of the controllers can not be guaranteed. A small delay such as this could cause destabilization of the network.
FIG. 1B
shows an example of a prior art node configuration. When messages of any type are received at the node at a particular interface, they first pass through switch
150
and then are forwarded to the layer
2
function of secondary controller
160
or of primary controller
165
depending on which controller is active. The layer
2
function of controllers
160
and
165
is also referred to as SSCOP (Service-Specific Connection-Oriented Protocol, set forth in the ITU (International Telecommunication Union) specifications Q.2110 (B-ISDN ATM adaptation layer-service Specific Connection Oriented Protocol (SSCOP) 07/94) and Q.2130 (B-ISDN signaling ATM adaptation layer-Service Specific coordination function for support of signaling at the user network interface (SSFC At UNI) 07/94), which provides a reliable data link transport function to layer
3
. The layer
3
function performs both signaling and call processing. Not shown is an intermediary buffer between layer
2
and layer
3
which is incorporated for sequencing and timing in some systems.
Because of the synchronization delay described above, SSCOP sessions may go down and communication with other nodes may be broken. When an SSCOP session goes down, ATM protocols require a status enquiry to be performed. The status enquiry resynchronizes all connections to the downed node with all its neighboring nodes. A node often has between 500,000 and 1,000,000 connections to resynchronize. If the synchronization does not occur quickly, all calls going to the downed node may be lost. It is desirable to avoid the status enquiry procedure because the procedure generates a flurry of status messages in the network—for example, if eight messages per connection are sent and the node maintains 100,000 connections, then 800,000 messages would be sent, which can cause a destabilization of the node.
For these reasons, switches have been typically operated on the same active primary card for extended periods of time. When the switchover is forced to occur, if a secondary controller is not operational, the primary controller has to be reset and the databases of the primary controller have to be rebuilt, causing hours of network outages.
SUMMARY OF THE INVENTION
What is disclosed is a method for switching from a primary network controller to a secondary network controller. A switchover request is received. Messages accepted by the node from the peer nodes are limited. An active link is maintained between the node and peer nodes. Finally, control is transferred from the primary controller to the secondary controller.
Other features and advantages of the present invention will be apparent from the accompanying drawings and from the detailed description that follows below.
REFERENCES:
patent: 5720029 (1998-02-01), Kern et al.
patent: 6170044 (2001-01-01), McLaughlin et al.
patent: 6272386 (2001-08-01), McLaughlin et al.
patent: 6335937 (2002-01-01), Chao et al.
patent: 6460149 (2002-10-01), Rowlands et al.
patent: 6490245 (2002-12-01), Burns et al.
ITU-T Recommendation Q.2110, B-ISDN -ATM Adaptation Layer -Service Specific Connection Oriented Protocol(SSCOP), pp. i-95, (Jul. 1994).
ITU-T Recommendation Q.2130, B-ISDN signaling ATM Adaptation Layer -Service Specific Coordination Function for Support of Signaling at the User-to-Network Interface (SSFC at UNI), pp. i-54 (Jul. 1994).
Karia Snehal
Medved Jan
LandOfFree
Method and apparatus for a graceful controller switchover does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for a graceful controller switchover, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for a graceful controller switchover will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3362450