Electrical computers and digital processing systems: multicomput – Distributed data processing – Client/server
Reexamination Certificate
1999-10-05
2001-09-04
Burgess, Glenton B. (Department: 2153)
Electrical computers and digital processing systems: multicomput
Distributed data processing
Client/server
C707S793000, C709S201000
Reexamination Certificate
active
06286032
ABSTRACT:
RELATED FIELD OF THE INVENTION
The invention generally relates to the field of communications, and more particularly, a communication network.
BACKGROUND OF THE INVENTION
A communication network, such as internet, comprised of a number of servers connected via a common network. Each server serves, i.e. manages, a number of processes that may be initiated by a client device using the network. The operations of such networks are commonly known by one of ordinary skilled in the art.
Referring to
FIG. 1
, a simplified block diagram of a communication network
100
is shown. Communication network
100
provides a common communication medium
101
for communications between a number of different client devices. Such devices may include computers, wire-line and wireless telephones. To manage the operations between such devices, common communication medium
101
connects servers
121
-
24
. Each device in order to use the network
100
communicates with a server selected from the available servers
121
-
24
. Network
100
may have many servers, however, only four servers are shown for simplicity. Each server may be serving a number of devices. Devices
171
-
73
and
174
are schematically shown, for example, where respectively each connecting to servers
121
and
124
at what commonly known as internet protocol (IP) addresses
131
-
34
respectively. Each device in network
100
would have a unique IP address identification.
Other devices may be connected to servers
121
-
24
, which are not shown.
A selected server starts a process on its processor for processing the needs of the client device. The needs may include communicating to another client device served by the same server or another server. A server may run multiple number of processes in order to serve each client device. Such processes at each server are shown by way of processes “1” through “m”, and collectively at each server as processes
161
-
64
.
A client device may need more than processes in order to use network
100
. Each process independently control a function or a feature of the communication requested by the client device. For example, a communication requested by a client device may include communications of voice, data and video where each requires an independent process. The selected server may also run another process for control and management of the related processes.
Each server requires having a memory device, such as memory devices
151
-
54
respectively associated with servers
121
-
24
, for storing record data associated with each process. The server that has been selected by the client device creates and maintains such record data, in effect, takes ownership of the record data.
The selected server also sends a copy of the record data associated with each process to other servers. Other servers after receiving the record data store the record data in their associated memory devices. As a result, the memory device of each server contains a copy of the record data that is similar to a copy of the record data at other servers.
The network
100
attempts to maintain consistency between record data stored at the memory device of each server through periodic communications between the server. Such periodic communications update the record data of each process every time the selected server, the record data owner, updates the record data. Updating such record data requires extensive communication traffic in the network
100
, which places a heavy burden of data traffic on common communication medium
101
.
To effectuate the record data management, each record data is required to have a key identifier in addition to an update sequence number. The update sequence number is incremented each time the data record is updated. The key identifier is shown by way of a set of symbols shown as S
N
m
. The “N” identifies the server that owns (selected) the record data, and “m” identifies the process within the selected server. For example, record data S
1
m
is a record data
180
for process “m” owned by server “1.” i.e. server
121
. A record data
185
is a record data for process “M” at server
124
. All record data are kept in a table format, for example, at memory table
186
at memory device
154
.
Each record data may have different parts. For example, record data
180
includes a character field
181
identifying the process, internet protocol (IP) address
182
identifying the client device, “unreliable datagram protocol” (UDP) ports address
183
identifying the logical address of the process at the client device where the process is running, and optional data
184
for different purposes. The record length may vary in different networks. Normally, the UDP address is 16 bits long, and IP address 32 bits. Longer or shorter record length is also possible.
In case a server shuts down or goes out of service for any reason, consistency of each record data stored at each memory device allows continuous and uninterrupted services for the client devices. Data owned by the out-of-service server is systematically adopted by one or more servers to allow continuous and uninterrupted services for the client devices. The network
100
requires a substantial communication bandwidth at common communication medium
101
to manage and accommodate updating such record data.
Therefore, there is a need for a method and apparatus in a communication network for updating and maintaining record data.
SUMMARY OF THE INVENTION
A communication network includes a first server and a second server, connected via a common network. First server owns a record data associated with at least one process running for serving a client device. The second server keeps a copy of the record data. The first server performs a hashing function over the record data to produce a Hash value. The hashing function includes concatenating UDP port information and at least a portion of an IP address to produce a string record A(e). The string record is squared to produce a squared record of A(e), and then the squared record of A(e) is truncated to at least a predetermined record length, such as 32 bits long. The second server similarly performs the same or similar Hashing function over the copy of the record data to produce a second Hash value. The first server sends the first Hash value to the second server for comparison. If the first Hash value fails to match the second Hash value, a latest copy of the record data is sent from the first server to the second server upon request.
REFERENCES:
patent: 4558413 (1985-12-01), Schmidt et al.
patent: 5765171 (1998-06-01), Gehani et al.
patent: 5829035 (1998-10-01), James et al.
patent: 5983241 (1999-11-01), Hoshino
patent: 5999947 (1999-12-01), Zollinger et al.
patent: 6003042 (2000-08-01), Melahn
patent: 6098078 (2000-08-01), Gehani et al.
patent: 6098079 (2000-08-01), Howard
Oberlander Lewis B.
Stewart Randall R.
Xie Qiaobing
Burgess Glenton B.
Edelman Bradley
Motorola Inc.
Williams Lalita P.
LandOfFree
Method and apparatus in a communication network for updating... 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 in a communication network for updating..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus in a communication network for updating... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2549769