PROVIDING ACCESS OVER AN IP NETWORK TO A SERVER APPLICATION PROGRAM

Field of Invention

The present invention relates to a method, apparatus or software for providing access by a client application program over an Internet Protocol (IP) network to a server application program instance.

Background of the Invention

In client/server computer systems, in order to provide increased server system bandwidth, a number of instances of a server application program may be provided. AU instances of the server application program share the same IP address and are arranged to accept communications on the same network port. A network port is an identification number used to pair incoming network data traffic to a given process or application running on the receiving computer. In other words, several instances of the server application program are active and waiting upon the IP address and port used by a given client application program. When a client application program connects to the IP address and port, one of the server instances is selected to establish a connection with the client application program.

During the subsequent interaction between the client application program and the selected instance of the server application program, data or other objects are likely to be created within the persistent storage for a selected instance of the server application program.

Further steps in the processing may be dependent on such data or objects. Therefore, in order to preserve such objects, the connection between the client application program and the selected instance of the server application program must be maintained.

One problem is that if the connection is removed or lost, the client application program will be unable to reconnect to the same instance of the server application program and elements

of the created data or objects may be lost. Furthermore, maintaining the connection requires significant network and server computer resources.

Summary of the Invention

An embodiment of the invention provides a method for providing access by a client application program over an Internet Protocol (IP) network to a server application program instance, the method comprising the steps of: a) providing IP network access for a plurality of instances of a server application program via a common port for a first IP address; b) enabling the connection of a client application program to a selected one of the plurality of server application program instances via the first IP address in response to a first request from the client application program; c) assigning a second IP address to the selected server application program instance for communication to the client application program, the second IP address being exclusive to the selected server application program instance; d) providing IP network access to the selected server application program instance via the second IP address; and e) enabling the reconnection of the client application program to the selected instance of the server application program via the second IP address.

The second IP address may be provided in response to a request from the selected instance of the server application program. The request from the selected instance of the server application program may be made in response to a corresponding request from the client application program. The second IP address may be selected from a predetermined set of unused IP addresses. The predetermined set of unused IP addresses may be determined from a set used IP address obtained from one or more relevant Domain Name Servers (DNSs). When the processing between the client application program and the selected instance of the server application program is completed, the second IP address may be freed for further use. The second IP address may be freed in response to a request from the client application program. The IP network and addresses may be IP version 6 (IPv6) compliant.

Another embodiment provides a method for providing access by a client application program over an Internet Protocol (IP) network to a server application program instance, the method comprising the steps of: a) connecting to a selected one of a plurality of instances of a server application program via a common port for a first IP address; b) requesting a second IP address from the selected server application program instance, the second IP address being exclusive to the selected server application program instance; c) in response to the receipt of the second IP address, closing the connection with the selected server application program instance; and d) reconnecting to the selected instance of the server application program via the second IP address.

A further embodiment provides a method for a server application program instance to provide a connection for a client application program over an Internet Protocol (IP) network, the method comprising the steps of: a) monitoring a port for a first IP address in common with a plurality of other instances of a server application program; b) in response to a first request from a client application program connecting to a the client application program via the first IP address to provide a first connection; c) in response to a second request from the client application program acquiring a second IP address exclusive to the server application program instance and communicating the second IP address to the client application program; d) in response to a third request from the client application program closing the first connection; e) monitoring a port for the second IP address; and f) in response to a fourth request from the client application program reconnecting to the client application program via the second IP address.

Another embodiment provides apparatus for providing access by a client application program over an Internet Protocol (IP) network to a server application program instance, the apparatus being operable to: provide IP network access for a plurality of instances of a server application program via a common port for a first IP address; enable the connection of a

client application program to a selected one of the plurality of server application program instances via the first IP address in response to a first request from the client application program; assign a second IP address to the selected server application program instance for communication to the client application program, the second IP address being exclusive to the selected server application program instance; provide IP network access to the selected server application program instance via the second IP address; and enable the reconnection of the client application program to the selected instance of the server application program via the second IP address.

A further embodiment provides apparatus for providing access by a client application program over an Internet Protocol (IP) network to a server application program instance, the apparatus being operable to: connect to a selected one of a plurality of instances of a server application program via a common port for a first IP address; request a second IP address from the selected server application program instance, the second IP address being exclusive to the selected server application program instance; in response to the receipt of the second

IP address, close the connection with the selected server application program instance; and reconnect to the selected instance of the server application program via the second IP address.

Another embodiment provides apparatus for a server application program instance to provide a connection for a client application program over an Internet Protocol (IP) network, the apparatus being operable to: monitor a port for a first IP address in common with a plurality of other instances of a server application program; in response to a first request from a client application program, connect to a the client application program via the first IP address to provide a first connection; in response to a second request from the client application program, acquire a second IP address exclusive to the server application program instance and communicate the second IP address to the client application program; in response to a third request from the client application program, close the first connection; monitor a port for the second IP address; and in response to a fourth request from the client application program, reconnect to the client application program via the second IP address.

A further embodiment provides a computer program product loadable into the internal memory of a digital computer, comprising software code portions for performing, when the product is run on a computer, a method for providing access by a client application program over an Internet Protocol (IP) network to a server application program instance, the method comprising the steps of: a) providing IP network access for a plurality of instances of a server application program via a common port for a first IP address; b) enabling the connection of a client application program to a selected one of the plurality of server application program instances via the first IP address in response to a first request from the client application program; c) assigning a second IP address to the selected server application program instance for communication to the client application program, the second IP address being exclusive to the selected server application program instance; d) providing IP network access to the selected server application program instance via the second IP address; and e) enabling the reconnection of the client application program to the selected instance of the server application program via the second IP address.

Another embodiment provides a computer program product loadable into the internal memory of a digital computer, comprising software code portions for providing, when the product is run on a computer, A apparatus for providing access by a client application program over an Internet Protocol (IP) network to a server application program instance, the apparatus being operable to: provide IP network access for a plurality of instances of a server application program via a common port for a first IP address; enable the connection of a client application program to a selected one of the plurality of server application program instances via the first IP address in response to a first request from the client application program; assign a second IP address to the selected server application program instance for communication to the client application program, the second IP address being exclusive to the selected server application program instance; provide IP network access to the selected server application program instance via the second IP address; and enable the reconnection of the client application program to the selected instance of the server application program via the second IP address.

Preferred embodiments of the invention use a feature of the IPv6 protocol which allows a given server to have many IP addresses. This feature can be used to allocate a unique temporary IP address by a server instance for a particular client in order to provide the ability for the client to re-establish access to that particular server instance without having to keep open the socket/connection to the server. Thus, the router and other physical resources required to keep such a socket/connection open may instead be made available for other users. At a later time, the client will be able to reopen contact with the specific server program instance previously used through use of the allocated IP address.

Brief Description of the Drawings

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which:

Figure 1 is a schematic illustration of a networked computer system;

Figure 2 is a schematic illustration of software components for a server computer in the computer system of Figure 1;

Figure 3 is a schematic illustration of software components for a client computer in the computer system figure 1 ;

Figure 4 is a table illustrating instructions for use between the client and server computers in the computer system of figure 1 ;

Figure 5 is a table illustrating instructions for use between software components of the server computer of figure 2;

Figure 6 is a flow chart illustrating processing performed by a client application program for the client computer of figure 3 ;

Figure 7 is a flow chart illustrating processing performed by an instance of a server application program for the server computer of figure 3; and

Figure 8 is a flow chart illustrating processing performed by an IP stack management program for the server computer of figure 2.

Detailed Description of Embodiments of the Invention

With reference to figure 1, a computer system 101 comprises a server computer 102 connected to a Wide Area Network (WAN) 103 along with a client computer 104 and a

Domain Name Server (DNS) 105. The server computer 102 and client computer 104 are arranged with software arranged to enable communication across the network 103 using the Transmission Control Protocol/Internet Protocol (TCP/IP) communications protocol. In the present embodiment, the TCP/IP version 6 (IPv6) protocol is used. IPv6 addresses are written as eight groups of four hexadecimal digits separated by colons and are typically composed of two logical parts: a 64-bit network prefix, and a 64-bit host part. The host part is commonly used to address an instance of a server application program and can by dynamically changed, that is, a server resource may, from time to time, effectively change its IPv6 address. The DNS 105 provides translation of network addresses in the form of Universal Resource Locators (URLs), such as www.ibm.com. into an IPv6 addresses, such as al29:e4e2:d5c6:b216:lc22:6ff2:57ab:9el2.

Figure 2 shows the software components of the server computer 102 in further detail. The server computer 102 is running an operating system 201, which, in the present embodiment, is the z/OS operating system from International Business Machines Corporation (IBM).

The operating system 201 is provided with an IP stack management program 202 in the form of the z/OS Communications Manager, arranged to provide an IP network interface that enables the operating system 201 to provide IP network access to resident programs. Three instances of a server application program 204, 205, 206 are running on the operating system 201, which, in the present embodiment, are instances of the IBM CICS Transaction Server

(CICS) application program. (IBM, z/OS and CICS are trademarks of International Business Machines Corporation in the United States, other countries, or both.) Each instance of the

server application program 204, 205, 206 has a common IP address and is arranged to listen on a common port provided by the IP stack management system 202 so as to provide high server availability. Thus, in response to a request from the client computer 104 for access to an instance of the server application program 204, 205, 206, one instance of the server application program 204, 205, 206 is selected at random and connected with the client computer 104. During the subsequent processing between the client computer 104 and the selected instance of the server application program 204, 205, 206, the data or other objects 207, 208, 209 that may be created are stored in the persistent storage for the selected instance of the server application program 204, 205, 206. Such data 207, 208, 209 defines an affinity or relationship between the client computer 104 and the selected instance of the server application program 204, 205, 206.

Figure 3 shows the software components of the client computer 104 in further detail.

The client computer 104 is running an operating system 301 providing a platform for a client application program 302 in the form of a web browser application program. The web browser 302 provides user access, via the IP network 103, to the services provided by instances of the server application program 204, 205, 206. The web browser 302 further comprises a plug-in program extension referred to herein as an affinity address (AA) client plug-in 303. The AA client plug-in 303 is arranged to provide the web browser 302 with functionality for requesting that the IP address for a given instance of the server application program 204, 205, 206 to which the web browser 302 is connected, be reassigned to a new IP address that is unique to the given instance of the server application program 204, 205, 206. This new IP address is, herein, termed an affinity address (AA). Preferably, the feature of the IPv6 protocol, which enables a given server to have multiple IPv6 addresses, is utilised. An IPv6 address is of the form a: b: c: d: w: x: y: z, where a, b, c, d, w, x and y are hex-as-character entries from 0000 to FFFF. In effect the a: b: c: d part of the IPv6 address is used for routers to direct an item through the network, and the w: x: y: z portion is used to identify the server which is communicating with the router. The server can request a unique temporary IPv6 address. Typically, the a: b: c: d: part of the affinity address of a server application program will be the same as that part of the first IP address of a server

application program, with the w: x: y: z part being changed to an address which is currently unallocated.

When the AA client plug- in 303 receives a requested affinity address it is arranged to close its current connection to the given instance of the server application program 204, 205, 206 and reconnect to the same instance via the received affinity address. This reconnection enables the relevant processing context, that is the data 207, 208, 209, to be preserved without the need to constantly maintain a connection between client application program 302 and given instance of the server application program 204, 205, 206.

Figure 4 shows the requests available to the AA client plug-in 303 by an Application Program Interface (API) provided with the instances of the server application program 204, 205, 206. The first request (Provide Affinity Address) is for the instance of the server application program 204, 205, 206 to provide a new affinity address, which, as described above, results in a new affinity address being provided to the AA client plug-in 303. The second request (Release Affinity Address) is for the instance of the server application program 204, 205, 206 to release a given affinity address. This results in the given instance of the server application program 204, 205, 206 being shut down and the affinity address released by the IP stack management program 202 for potential further use. In other words, the affinity addresses provided in this manner may only be temporarily assigned to a given instance of the server application program 204, 205, 206.

In the present embodiment, the API between the AA client plug-in 303 and the server application program 204, 205, 206 is provided via an extension to the Hypertext Transfer Protocol (HTTP) version 1.1 Options command. The Options command is used to transport the Provide and Release requests and responses between the AA client plug-in 303 and respective instances of the server application program 204, 205, 206 as HTTP message headers. Thus for example, in order to request an affinity address, the AA client plug-in 303 sends an HTTP message including a header named Provide Affinity Address. The response from a respective instance of the server application program 204, 205, 206 will be an HTTP message including the header name Affinity_Address: 1 :2:3:4:5:6:7:8, where 1 :2:3:4:5:6:7:8 is the requested affinity address.

Returning to figure 2, each instance of the server application program 204, 205, 206 further comprises a plug-in program extension referred to herein as an affinity address (AA) server plug-in 210, 211, 212. In addition, the communications manager 202 further comprises a plug-in program extension referred to herein as an affinity address (AA) provider plug-in 213. Each instance of the AA server plug-in 210, 211, 212 is arranged to provide functionality to a given instance of the server application program 204, 205, 206 for routing requests between the AA client plug-in 303 and the AA provider plug-in 213. The AA provider plug- in 213 is arranged to provide functionality to the communications manager 202 for identifying and managing a set of affinity addresses 214 and for supplying unique affinity addresses on request. The management of the affinity addresses comprises obtaining a list of all relevant used IP addresses from the DNS 105 and deducing the set of available addresses, that is, the set of possible affinity addresses. Furthermore, in response to requests to provide or release affinity addresses, the AA provider plug-in 213 is arranged to update the availability status of the affinity address respectively both in the set of affinity addresses 214 and with the DNS 105.

Figure 5 shows the requests provided to each instance of the AA server plug-in 210, 211, 212 by the API provided with the AA provider plug-in 213. The first request (Get_Affinity_Address) is issued by the AA server plug-in 210, 211, 212 to the AA provider plug-in 213 in response to a Provide Affinity Address request received from the AA client plug-in 303. In response, the AA provider plug-in 213 selects an affinity address, marks the selected address as unavailable and returns the selected address to the AA server plug-in 210, 211, 212. The selected affinity address is then passed to the requesting AA client plug- in 303 and also assigned to the given instance of the server application program 204, 205, 206 to enable the reconnection of the client application program 302 and the instance of the server application program 204, 205, 206 via the affinity address. The second request (Free_Affinity_Address) in the API of figure 5 is issued by the AA server plug-in 210, 211, 212 to the AA provider plug-in 213 in response to a Release_Affinity_ Address request received from the AA client plug-in 303. In response, the AA provider plug-in 213 marks the freed affinity address as available. The third and fourth requests are provided as part of a management interface for the AA provider plug-in 213 and respectively enable all affinity

address in use by a given server and all known affinity addresses in the IP stack to be discovered.

The processing performed by the AA client plug- in 303 in combination with the client application program 302 will now be described in further detail with reference to the flow chart of figure 6. At step 601, processing of the AA client plug-in 303 is initiated in response to the initialisation of the client application program 302 and processing moves to step 602. At step 602, a connection is made to a randomly selected instance of the server application program (Server Instance 1 in figure 2) 204, over the IP network 103 and processing between the client and server programs 302, 204 is initiated. Processing then moves to step

603 where a request for the provision of an affinity address is sent to the selected instance of the server application program 204 and processing moves to step 604. At step 604, the requested affinity address is awaited and once received processing moves to step 605. At step 605, the processing between the client and server programs 302, 204 is temporarily suspended and processing moves to step 606. At step 606, the current connection with selected instance of the server application program 204 is closed, thereby freeing up all the resources for the communication, and processing moves to step 607. At step 607, a new connection is made to the same selected instance of the server application program 204 via the received affinity address and processing moves to step 608. At step 608, the processing between the client and server programs 302, 204 is resumed and processing moves to step

609. At step 609, the completion of the processing between the client and server programs 302, 204 is awaited and once completed processing moves to step 610. At step 610, a request to release the received affinity address is sent to the server application program 204 and processing returns to step 602 to await the initiation of a further connection to a server.

It is to be noted that in prior art situations, rather than closing the connection in step 606, the current connection with a selected instance of the server application program 205 must be kept active, with the communication resources continuing to be allocated and thus unavailable for use by others.

The processing performed by an instance of the AA server plug-in 210 in combination with an associated instance of the server application program 204, 205, 206 will now be described

in further detail with reference to the flow chart of figure 7. At step 701, the processing of an instance of the AA server plug-in 210 is initiated in response to the initialisation of the associated instance of the server application program 204 and processing moves to step 702. At step 702, a requested connection from the client application program 302 is received and accepted and processing moves to step 703. At step 703, the processing between the client and server applications 302, 204 is performed and continues until either an affinity address related request is received from the client application program 302 or until processing is terminated at which point processing moves to step 704. At step 704, if an affinity address related request has been received processing continues to step 705. If at step 705 the request is identified as a request for the provision of an affinity address processing moves to step

706. At step 706, a request for a new affinity address is sent to the AA provider plug-in 213 and processing moves to step 707 to await receipt of the new affinity address. Once the affinity address is received processing moves to step 708. At step 708, the received affinity address is passed back to the requesting client application program 302 and processing moves to step 709. At step 709, the new affinity address is assigned to the current instance of the server application program 204, which then listens on the affinity address for a reconnection request from the client application program 302 and once received processing moves to step 710 where the reconnection is established and processing between the client and server application programs 302, 204 is resumed. When processing between the client and server application programs 302, 204 is terminated processing moves to step 711 and ends for this instance of the server application program 204. If at step 704 the received request is a request to terminate the processing between the client and server application programs 302, 204 then processing moves to step 711 and ends as described above. If at step 705 the received affinity address related request is a request to release a current affinity address then processing moves to step 712 where the appropriate request is sent to the AA provider plug-in 213 and processing then moves to step 711 and ends as described above.

The processing performed by the AA provider plug-in 213 in combination with the IP stack management program 202 will now be described in further detail with reference to the flow chart of figure 8. At step 801, the processing of the AA provider plug-in 213 is initiated in response to the initialisation of the processing of the IP stack management program 202 and processing moves to step 802. At step 802, a list of all currently used IPv6 addresses is

collected from the DNS 105 and processing moves to step 803. At step 803, a list of unused IPv6 address is deduced from the list of currently used IPv6 addresses and processing moves to step 804. At step 804, the list of unused IPv6 addresses is saved as the set of possible affinity addresses 214 and processing moves to step 805. At step 805, processing waits for an affinity address request from an instance of a server application program 204 and once received processing moves to step 806. If at step 806, the request is to provide a new affinity address then processing moves to step 807. At step 807, an address is selected from the set of affinity addresses 214 and marked as in use and processing moves to step 808. At step 808, the DNS 105 is informed that the affinity address is in use and processing moves to step 809. At step 809, the selected affinity address is returned to the requesting instance of the server application program 204 and processing moves to step 805 to await the receipt of a further affinity address related request. If at step 806, the request is for the release of an in-use affinity address then processing moves to step 810. At step 810, the DNS 105 is updated accordingly and processing moves to step 811. At step 811, the freed affinity address is marked as available in the set of affinity addresses 214 and processing returns to step 805 and proceeds as described above.

As will be understood by those skilled in the art, while embodiments of the invention are described with respect to IPv6, other embodiments may use other protocols with suitable mechanisms for dynamically assigning unique network addresses to a given instance of an application program. Furthermore, while embodiments of the invention are described with respect to TCP, other network transport protocols suitable protocols such as User Datagram Protocol (UDP) may be substituted or used in combination with TCP in some embodiments.

In the above embodiments, the selection of and instance of a server application program for connection to a requesting client application program is made at random. As will be understood by those skilled it the art, other criteria such as server workload balancing or failover mechanisms may contribute at least in part to the selection of a given instance of the server application program.

In the above embodiment, functionality is added to the client and server application programs and the IP stack management program using a respective program function

extension or plug-in. As will be understood by those skilled in the art, the functions provided by any of the plug-ins described above may be provided instead by any other suitable means such as by adding to or modifying the code for the original client or server application programs or the IP stack management program. The functions of the server and provider plug-ins described herein may also be provided by or incorporated into other suitable programs within the respective systems, for example by the addition of appropriate affinity address server and provider routines.

In the above embodiment, the DNS provides a list of used IP addresses from which the AA provider plug-in derives and then manages the set of available affinity addresses (AAs). In other embodiments, the management of the pool of AAs may be performed solely by the DNS. In another embodiment, the management of the pool of AAs may be shared between the DNS and the AA provider plug-in.

In a preferred embodiment, given an IPv6 address with the form a:b:c:d:e:f:g:h where the a:b:c:d part is fixed, elements of the non-fixed part are reserved for indicating that an address is an affinity address. In this embodiment, the DNS need not be involved in the selection or management of affinity addresses.

It will be understood by those skilled in the art that the apparatus that embodies a part or all of the present invention may be a general purpose device having software arranged to provide a part or all of an embodiment of the invention. The device could be a single device or a group of devices and the software could be a single program or a set of programs. Furthermore, any or all of the software used to implement the invention can be communicated via any suitable transmission or storage means so that the software can be loaded onto one or more devices.

While the present invention has been illustrated by the description of the embodiments thereof, and while the embodiments have been described in considerable detail, it is not the intention of the applicant to restrict or in any way limit the scope of the appended claims to such detail. Additional advantages and modifications will readily appear to those skilled in

the art. Therefore, the invention in its broader aspects is not limited to the specific details representative apparatus and method, and illustrative examples shown and described.

For the avoidance of doubt, the term "comprising", as used herein throughout the description and claims is not to be construed as meaning "consisting only of.