Method and apparatus of improving network performance and network availability in a client-server network by transparently replicating a network service

• US 6,097,882 A
• Filed: 06/30/1995
• Issued: 08/01/2000
• Est. Priority Date: 06/30/1995
• Status: Expired due to Term

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Digital Equipment Corporation (HP Inc.)
• Inventors: Mogul, Jeffrey C.
• Primary Examiner(s): Rinehart, Mark H.
• Application Number:
  US08/497,095
• Time in Patent Office: 1,859 Days
• Field of Search: 395/200.05, 395/200.06, 395/200.1, 395/200.12, 395/200.13, 395/200.15, 395/200.18, 395/200.01, 395/200.02, 395/200.38, 395/200.55 , 395/200.5, 395/200.57, 395/200.63, 395/200.68, 395/200.76, 395/200.3, 395/200.79
• US Class Current: 709/201
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/1017 based on a round robin mechanism
• CPC Class Codes: H04L 67/1034 Reaction to server failures by a load balancer
• CPC Class Codes: H04L 69/16 Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
• CPC Class Codes: H04L 69/163 In-band adaptation of TCP data exchange; In-band control procedures

Abstract

A client-server network including a number of client computer systems, each of the client computer systems having a network interface, a number of server computer systems, each of the server computer systems having a network interface, and a replicator system connecting the client computer systems to the server computer systems, the replicator system transparently processing a number of requests from the client systems to a number of services resident in the server systems.

Claims

1. A client-server network comprising:

a plurality of client computer systems, each of the client computer systems having a network interface;

a plurality of server computer systems, each of the server computer systems having a network interface; and

a replicator system connecting the plurality of client computer systems to the plurality of server computer systems, the replicator system transparently processing a plurality of requests from the client systems to a plurality of services resident in the server systems.

2. The client-server network according to claim 1 wherein the replicator system comprises:

means for receiving a connection request and a plurality of subsequent packets of data associated with the connection request from a first client computer system to a first server computer system;

means for processing the connection request according to a plurality of policy parameters;

means for storing the connection request;

means for forwarding the connection request to one of the plurality of server computer systems;

means for receiving a reply from the one of the plurality of server computer systems in response to the connection request;

means for processing the reply;

means for storing the reply; and

means for forwarding the reply to the first client computer system.

3. The client-server network according to claim 2 wherein the means for processing the connection request and the subsequent data packets comprises;

means for checking a first destination address of the connection request;

means for relating the first destination address to a second destination address according to a set of policy parameters;

means for storing the first destination address and the second destination address in a table in the replicator system;

means for detecting when the connection is aborted or closed.

4. The client-server network according to claim 3 wherein the set of policy parameters are selected form one of the following:

a performance variable of the first server computer system and a performance variable of the second sever computer system;

a plurality of parameters representing past operations, such plurality including round robin scheduling and affinity scheduling;

a failover parameter, the failover parameter indicating a system status of the first server computer system and the second server computer system; and

a load balance system parameter.

5. The client-server network according to claim 1 wherein the replicator system comprises:

a first replicator system;

a second replicator system;

a connection between the first replicator system and the second replicator system, the connection used to coordinate processing in the first replicator system with processing in the second replicator system.

6. An arrangement for interfacing a plurality of client systems and a plurality of server systems in a computer network comprising:

a first network interface to receive a connection request from a first client system, the connection request containing an origination address indicating the first client system and a destination address indicating a first server system;

a memory to process the connection request and a plurality of subsequent data packets associated with the connection request and relate the destination address to one of the plurality of server system according to a set of policy parameters;

a storage subsystem to store the connection request and the relationship;

a second network interface to send the connection request to the one of the plurality of server systems;

a third network interface to receive an acknowledgment packet from the one of the plurality of server systems to the first client system;

the memory processing the acknowledgment packet according to the relationship; and

a fourth network interface to send the acknowledgment to the first client system.

7. The arrangement according to claim 6 wherein the first network interface and the fourth network interface are a single network interface.

8. The arrangement according to claim 6 wherein the second network interface and the third network interface are a single network interface.

9. The arrangement according to claim 6 wherein the first network interface, the second network interface, the third network interface, and the fourth network interface is a single network interface.

10. The arrangement according to claim 6 wherein the set of policy parameters is selected from one of the following:

a performance variable of the first server system and a performance variable of a second server system;

a failover parameter, the failover parameter indicating a system status of the first server system and the second server system; and

a load balance system parameter.

11. The arrangement according to claim 6 wherein the computer network uses the TCP/IP protocol.

12. The arrangement according to claim 6 wherein the computer network uses the UDP/IP protocol.

13. The arrangement according to claim 6 wherein the memory is non-volatile random access memory (NVRAM).

14. A method of improving performance and availability in network having a plurality of clients and a plurality of servers comprising the steps of:

providing a connection request from a first client to a first server;

intercepting the connection request;

processing the connection request according to a set of policy parameters to produce a relationship of addresses;

storing the relationship of addresses;

forwarding the connection request to one of the plurality of servers according to the relationship of addresses;

receiving an acknowledgment packet from the one of the plurality of servers;

processing the acknowledgment packet according to the stored relationship of addresses;

forwarding the acknowledgment packet to the first client according to the relationship of addresses; and

Intercepting a plurality of subsequent data packets associated with the connection request;

repeating the steps of processing, forwarding, receiving, processing, and forwarding for the plurality of subsequent data packets.

15. The method of improving performance and availability in a network according to claim 14 where the set of policy parameters is selected from one of the following:

a performance variable of a first server and a performance variable of a second server;

a failover parameter, the failover parameter indicating a system status of the first server and the second server;

a plurality of parameters representing past operations, such plurality including round robin scheduling and affinity scheduling; and

a load balance system parameter.