AU709604B2

AU709604B2 - Method of linking processor units in exchange control

Info

Publication number: AU709604B2
Application number: AU65755/96A
Authority: AU
Inventors: Detlef Daase; Karl-Heinz Legat
Original assignee: Alcatel NV
Current assignee: Alcatel Lucent NV
Priority date: 1995-09-30
Filing date: 1996-09-20
Publication date: 1999-09-02
Anticipated expiration: 2016-09-20
Also published as: CN1130654C; AU6575596A; DE19536649A1; CN1151674A; EP0766488B1; ATE325508T1; DE59611342D1; EP0766488A3; JPH09231243A; EP0766488A2; CA2186777A1

Abstract

The coupling method involves two data processing units (PU1,PU2) which have different operating systems (OS1,OS2). The method allows communication between their respective databanks (DBMS1,DBMS2). The databanks are coupled via a replication mechanism between their databank managers (DBM1,DBM2). The processors (P1-P3) for the first processing unit employ access to semi-permanent variables of the processors (P4-P6) for the second data processing unit. Preferably the data processing units have different, problem-definition-adapted operating systems and databanks.

Description

This invention relates to a method for linking data processing units which each operate with an independent operating system.

For the linking of sequences or processes in a computer system it is necessary for these to communicate with each other. By sequence we mean here the execution of a program, and the execution of a program is considered a process if this execution is managed by an operating system (see also Chapter 1.2 Processes, page 15 of the book "Betriebssysteme: eine Einf(hrung" [Operating Systems: An Introduction] by H.-J.

Siegert, Oldenburg Verlag, 1991).

Methods for making possible the linking of processes managed by the same operating system are, for example, described in the chapter "Prozesskommunikation" [Process Communication] of the book "Betriebssysteme: eine Einfoihrung" [Operating Systems: An Introduction] by Siegert, Oldenburg Verlag, 1991, on pages 81-96.

There the communication between the processes, also known as interprocess communication, is achieved on the one hand by the exchange of messages by means of operating system services. On the other hand, it is also possible for the communication to take place without the knowledge or support of the operating system, by means of the exchange of messages via common storage areas provided by the operating system. The processes taking part in the communication can here also execute on different computers, if an operating system is being considered which is distributed over several computers. These methods are, however, only usable for processes managed by the same operating system, and not for communication between processes which are managed by different operating systems.

Furthermore, it is well known to manage data centrally for a number of processes by means of a database. The database makes it possible for the processes to access the data in accordance with the data model implemented in the database. It is here also possible to access such a database from another computer.

9999 The invention is based on a control unit for an ISDN communication system (ISDN Integrated Services Digital Network) which is described in the German Patent DE 41 04 365 C1 and in which processes managed by different operating systems communicate with each other.

This control unit has two operating systems at its disposal which operate on the same computer. Here we are dealing on the one hand with a subscriber operating 3 system for subscriber- specific data processing, and on the other hand with a real-time operating system for communication tasks. These operating systems are linked in such a way that the data or data application managed by the real-time operating system is made available for access by the processes of the subscriber operating system, and on the other hand the data applications conducted under the subscriber operating system can be taken over by processes of the real-time operating system. The communication between processes which are managed by different operating systems is achieved here by an exchange of messages via a common memory segment. This common memory segment, which is described as an Intercommunication Data Segment, is defined by two equal data segment scripts which thus make this data segment usable by sequences and processes of both operating systems.

This method of linking has the disadvantage that at least one computer must be operating in parallel with both operating systems in order to make access possible to the common data segment by both operating systems.

Furthermore, it is a-disadvantage that every data transfer between processes of different operating systems requires costly, precisely-matched transfer procedures in all "the processes taking part. This involves a large implementation cost, and complex interdependencies between control programs working with different operating systems.

An object of the present invention is now to enable the construction of a control 20 unit of data processing units which operate with independent operating systems.

A data processing unit is here understood to be a logical unit which consists of an operating system and the control programs working on this operating system, as well as the hardware platform it is based on. In general for a data processing unit this will be a computer, together with the associated peripheral components, which operates with an independent operating system. It can, however, also consist of a *computer system with a distributed operating system, or even a logical or functional part of a computer when the computer has several independent operating systems.

According to a first aspect of the invention, there is provided a method for linking data processing units which each operate with an independent operating system, wherein in each of the data processing units an independent database is provided which manages at least a part of the data of the respective data processing unit according to a corresponding data model, that a first database of a first data processing unit and a second database of a second data processing unit are linked in such a way that the first database accesses a data record of the second data processing unit which is managed by the second database, and the first database makes this data processable according to its data model for processes of the first data processing unit and that, for cooperation between the first and second data processing units, the processes of the first data processing unit, by accessing the first database, access data from the data record of the second data processing unit.

According to a second aspect of the invention, there is provided a method for controlling an exchange, wherein two or more data processing units working with different operating systems cooperate for the performance of switching functions, that in each of the data processing units an independent database is provided which manages at least a part of the data of the respective data processing unit according to a respective data model, that a first database of a first data processing unit and a second database of a second data processing unit are linked in such a way that the first database accesses a data record of the second data processing unit which is managed by the second database, and makes this data visible according to its data model for processes of the first data processing unit and that, for cooperation between the first and second data processing units, the processes of the first data processing ,.unit, by accessing the first database, access data from the data record of the second data processing unit.

According to a third aspect of the invention, there is provided a first data processing unit with at least one computer system as a hardware platform, with an independent operating system and with a number of control programs at least one of awhich is so designed that, in its operation, it cooperates with at least one data processing unit which in turn is provided with an independent operating system, *wherein the first data processing unit is provided with a first database for managing the data in accordance with a first data model, that the first database is linked to a second database of the second data processing unit which is so designed that it manages at least a part of the data of the second data processing unit in accordance with a second data model, that the first database is so designed that it accesses a data record of the second data processing unit where the data is managed by the second database and made visible to the processes of the first data processing unit in accordance with the first data model, and that the control program is so designed that during its execution, in order to cooperate with the second data processing unit, it accesses data of the data record of the second data processing unit by accessing the first database.

According to a fourth aspect of the invention, there is provided a control unit with at least one computer system as a hardware platform and with at least a first and a second data processing unit, where the first and second data processing unit are each respectively provided with an independent operating system and with a number of control programs of which at least one control program of the first data processing unit is so designed that in its execution it cooperates with the second data processing unit, wherein the first and the second data processing unit are provided with a first and a second database which are respectively so designed that they respectively manage at least a part of the data of the respective data processing unit in accordance with a respective data model, that the first database is linked to the second database, that the first database is so designed that it accesses a data record .9of the second data processing unit, whose data is managed by the second database, and makes it visible in accordance with its data model for processes of the first data processing unit, and that the control program is so designed that during its execution, for cooperation with the second data processing unit, by accessing the first database, it accesses data of the data record of the second data processing unit.

The basic idea of the invention is to manage at least a part of the data of each .9 9. 9 of the data processing units respectively in a database, and to link these databases in such a way that the data of the respective other data processing unit is made visible in *accordance with the data model of the respective database for processes of the respective data processing unit. For the processes it then appears as if the data is managed by the database of its own data processing unit. Via this linking of the databases, the data of another data processing unit is then accessed. By manipulating or interpreting this data, a deep intervention becomes possible into these processes which operate with this data and are managed by another operating system, and thus a deeper intervention becomes possible in the other data processing unit.

It is advantageous to use this basic idea for the construction of control units for exchanges. The tasks for such control units place very diverse demands on the operating system which is used. There are thus advantages if such a control unit is constructed from several data processing units which operate with different operating systems and are linked for cooperation in accordance with the above basic idea.

The resulting advantages here consist especially of the reduction of the software implementation and testing costs, since operating systems, databases and programming languages can be used which are matched to the respective types of task.

Advantageous design examples of the invention can seen in the subsidiary claims.

The invention has the advantage that diverse possibilities for views and intervention into sequences or processes of another operating system become available, where these views and interventions do not already have to be defined precisely in advance with mutually synchronised transfer routines. Because of this, the interdependencies are reduced which must be taken into account in the development of corresponding control programs. Even a later-required extension of the views or interventions into another data processing unit therefore entails only a small cost.

Altogether, the development, test, and maintenance costs are reduced for the 0600 construction of control units which consist of several cooperating data processing units.

An additional advantage of the invention is that it makes possible the separation of the development cycles of different cooperating data processing units.

This advantage results from the fact that, because of the linking in accordance with the invention, a further development of the control programs for a data processing unit only has a small effect on the control programs of other data processing units which cooperate with this one.

25 In order that the invention may be readily carried into effect, embodiments thereof will now be described in relation to the accompanying drawings, in which: Figure 1 shows a block diagram of an exchange according to the invention, with a control unit according to the invention.

Figure 2 shows a functional representation of the control unit of Figure 1 according to the invention, for a first design example.

Figure 3a and Figure 3b show functional representations of a linking of two data processing units of the control unit of Figure 2 according to the invention.

Figure 4 shows a functional representation of the control unit of Figure 1 according to the invention, for a second design example.

Figure 5 shows a functional representation of the control unit of Figure 1 according to the invention, for a third design example.

In the first design example, the application of the method of the invention is described for the control of an exchange in an exchange according to the invention, whose control unit consists of two data processing units according to the invention which, by means of the method of the invention, are linked together for the linking of data processing units.

Figure 1 shows an exchange EX with a control unit CONTR and a switching network DSN. The control unit CONTR and the switching network DSN exchange data with each other.

The control unit CONTR serves for the control of all switching functions and service features provided by the exchange

EX.

The switching network DSN is a conventional digital switching network for exchanges. It is controlled by the control unit CONTR and serves both for the switching of connections for subscribers of the exchange EX and as a communication medium for the internal communication of the control unit CONTR.

It is also possible to use an independent communication network for this internal communication. Such a communication network can also be a network with a distributed switching process, for example a LAN (Local Area Network) with an Ethernet or Token Access protocol.

The control unit CONTR has four independent computer systems CS1 to CS4 which communicate with each other via the switching network DSN.

Each of the computer systems CS1 to CS4 consists of a computer and peripheral components connected to it which are required for the provision of the respective functions, and for communication with the remaining ones of the computer systems CS1 to CS4.

The computer system CS1 consists of a computer which has been equipped with one or more processors, an internal computer bus, an internal memory, at least one mass-storage device such as e.g. a hard disk drive, and with components for 8 input and output of data. In addition, the computer system CS1 is provided with peripheral components which facilitate man/machine communication and the exchange of data via the switching network DSN.

The computer systems CS2 to CS4 are systems designed to provide telecommunication functions and their construction can, for example, be seen in the article "Hardware Struktur" [Hardware Structure] pages 135 to 147 of the journal Elektrisches Nachrichtenwesen [Electrical Communication Systems], Vol. 58, No. 2/3, 1981, or in other articles of this volume.

Each of the computer systems CS1 to CS4 forms a hardware platform for a number of control programs and for an operating system which, in particular, supports and manages the execution of the control programs. The exact definition of an operating system can, for example, be taken from the standard DIN 44300. The control programs are stored in the respective computer systems and their execution determines the function of the respective computer system.

The computer system CS1 operates with an operating system OS1 and the computer systems CS2 to CS4 with a distributed operating system OS2, where a part of the operating system operates on each of the computer system CS2 to CS4.

The execution of the control programs on the computer system CS1 is thus managed by the operating system OS1, and the execution of the control programs of the computer systems CS2 to CS4 is managed in common by the operating system OS2. The computer system CS1 together with the associated control programs and the operating system OS1 thus forms a data processing unit PU1, and the computer to systems CS2 to CS4 with the associated control programs and the distributed operating system OS2 form another data processing unit PU2. The operating systems OS1 and OS2 are independent operating systems, i.e. they manage the execution of the respective control programs independently and are not operating-system parts of a higher-order distributed operating system.

It is also possible for the data processing unit PU2 to consist only of one computer system. Furthermore, it is possible that the data processing units PU1 and PU2 are logical data processing units whose processes, although respectively managed by an independent operating system, nevertheless share a common hardware platform. The computer system would thus operate in parallel with both operating systems OS1 and OS2.

The functional construction and the joint operation of the data processing units PU1 and PU2 will now be explained with the aid of Figure 2.

Figure 2 shows the control unit CONTR with the data processing units PU1 and PU2. The data processing unit PU1 has three processes P1 to P3, a database DBMS1 and the operating system OS1. The data processing unit PU2 has three processes P4 to P6, a database DBMS2 and the operating system OS2. The processes P1 to P3 communicate with each other and with the database DBMS1. The processes P4 to P6 communicate with each other and with the database DBMS2. The databases DBMS1 and DBMS2 are linked together.

The processes P1 to P3 respectively represent the execution of a control program by the computer system CS1, e.g. the achievement of a particular task by the computer system CS1. With the processes P1 to P3 we are here dealing with such processes which provide functions in the framework of the switching tasks of the control unit CONTR. The number of three processes P1 to P3 is here only chosen as an example and usually changes continually depending on the state of the data processing unit PU1.

Similar considerations apply to the processes P4 to P6.

The operating system OS1 manages the processes P1 to P3 and, for this purpose, in particular determines the sequence of their execution, starts them, determines their state and provides standard services to them which, in particular, ~carry out the transfer of data between processes, and between processes and units of so the computer system CS1. Furthermore during the simultaneous execution of several processes, the operating system OS1 coordinates and synchronises their access to common operating resources such as, for example, the computer CPU, memory or 25 ancillaries.

059.

Similar considerations apply to the processes P4 to P6.

In detail, the operating system OS2 is a real-time operating system which is specially designed for the provision of functions which are required for the control of an exchange. The operating system OS1, on the other hand, is a non-real-time operating system such as is customarily used in general-purpose computers. An example of such an operating system is the operating system known as UNIX.

It is also possible for the operating systems OS1 and OS2 to be operating systems of the same kind.

The database DBMS1 manages a part of the data of the processing unit PU1, with the respective data model which is implemented. This database is an independent database, i.e. it is not part of a higher-order distributed database. With this data we are, for example, concerned with semipermanent variables or with databases which are used by the control programs of the data processing unit PU1. Access to this data is achieved here by having the processes P1 to P3 request or manipulate the data from the database DBMS1 by means of special messages.

From a logical point of view, the database DBMS1 thus represents one or more specialised processes which manage data and on request make it available to the processes P1 to P3, or manipulate it according to their requirement. The communication between the processes P1 to P3 and the database DBMS1 is here supported by services of the operating system OS1 which in particular makes possible the distribution of the processes P1 to P3 and the database DBMS1 to different computer systems which operate with a common, distributed operating system.

Similar considerations apply to the database DBMS2.

With the databases DBMS1 and DBMS2 we are dealing with different databases. These databases are chosen according to the respective system platforms, i.e. the computer platform as well as the operating system, and according to the 20 functions to be provided by the data processing units PU1 and PU2. Thus the database DBMS1 is, for example, a conventional relational database such as is described in the book "The SQL Standard", C.J. Date, 1993, Addison Wesley. Such a database is, for example, the database sold under the name ORAKEL [ORACLE]. The database DBMS2, on the contrary, can for example be a special, distributed database optimised for the provision of switching functions.

It is, however, also possible for the two databases DBMS1 and DBMS2 to be the o same kind of database.

In more detail, the databases DBMS1 and DBMS2 each have an access interface INT1 or INT2, a database manager DBM1 or DBM2, and a database store 30 DB1 or DB2.

The processes P1 to P3 access the database DBMS1 via the access interface 11 INTl. The processes P4 to P6 access the database DBMS2 via the access interface INT2. The database managers DBM1 and DBM2 access the database stores DB1 and DB2 and communicate with each other.

The access interfaces INT1 and INT2 provide the interfaces between the databases DBMS1 and DBMS2 and the processes P1 to P3 and the processes P4 to P6 which, in this connection, are also referred to as application processes. For example, they thus support a database access language by means of which access to the associated database becomes possible. Such database access languages are, for example, the database access languages SQL (Structured Query Language) or NDL (Network Database Language). Special proprietary database store access languages or access protocols can also be used.

The database store DB1 represents a storage medium in which the data managed by the database DBMS1 is physically stored. This storage medium consists of one or more mass storage units of the computer system CS1, for example hard disk drives. Corresponding considerations apply to the database store DB2 which manages the data of the database DBMS2.

It is possible here for the database store to be distributed over the computer systems CS2 to CS4, i.e. that a part of the data is stored in the storage systems of the computer systems CS2, CS3 or CS4 respectively.

The database managers DBM1 and DBM2 carry out the actual management •:j0 functions for the data of the databases DBMS1 and DBMS2 which is stored in the database stores DB1 and DB2. They determine the possible access procedures for the data and manage the data model according to which the access to the data is carried out. By means of their knowledge of this data model they carry out the actual access operations for the data stored in the respective database stores.

Furthermore, the databases DBMS1 and DBMS2 are linked via the database managers DBM1 and DBM2. Because of this linking, a selected data record of the data processing unit PU2 is made visible to, that is capable of being processed by, the processes P1 to P3, in accordance with the data model managed by the database manager DBM1. This means that the processes P1 to P3 can access this data record by *30 means of the access procedures of database DBMS1 and according to its data model.

Accessing this data record is no different for it from accessing data of its own data processing unit PU1. Correspondingly, the converse applies for the processes P4 to P6.

The switching functions are provided by the control unit CONTR as follows: The data processing unit PU2 is responsible for carrying out standard switching tasks or basic switching tasks in real time. The individual tasks are here provided by the processes P4 to P6.

The data processing unit PU1 is responsible for the provision of more complex switching functions for which there is no high real-time requirement. The tasks of such functions, for example, consist of the provision of ISDN service features, subscriber service features, Centrex services, or of the metering of individual calls or the applications or services for mobile phone systems or intelligent networks. These tasks are provided by the processes P1 to P3.

The data which is needed by the processes P1 to P6 for carrying out their tasks is managed by the databases DBMS1 or DBMS2. Such data consists, for example, of semipermanent variables of the control programs whose execution is represented by the processes P1 to P6, subscriber data, data concerning the state of the switching network DSN, data about the telecommunication services to be provided, or data about traffic loading of the exchange EX. Through the linking of the databases DBMS1 and DBMS2, this data is available for all the processes P1 to P6 and they can access this data using their own accustomed access procedures according to the data model of their database, independently of which of the two databases respectively manages the data. Furthermore, the databases DBMS1 and DBMS2 also manage, and via the linking make visible respectively for processes of the other database, data which determines the state of the processes, or whose changes alter the execution or the result of the processes. The extent to which data is made visible to the respective other data processing unit by the linking, here determines the depth of the mutual intervention possibilities.

It is also possible that in the databases DBMS1 and DBMS2 additional data of S the data processing units PU1 and PU2 is managed. Such data can, for example, also be data that is used by the system processes of the operating systems OS1 or OS2 and affects their execution. It could also include parameters of the operating system 30 kernel, so that a direct influence is possible on the control of the execution of the respective other operating system.

13 It is furthermore possible that the data of the other data processing unit which is made visible by the linking is mapped onto an object-oriented information model, and thus the cooperation of the two data processing units is realised by means of an object-oriented module.

The linking of the two databases DBMS1 and DBMS2 by means of a replication mechanism will now be explained in more detail by means of Figure 3a and Figure 3b.

Figure 3a shows the database manager DBM1 and the data processing unit PU2 which exchange messages with each other. The database manager DBM1 has several database processes PDB, a converter MAP, a data record DATMOD and a communication unit KOM. The communication unit KOM exchanges messages with the data processing unit PU2 and receives messages from the database processes PDB via the converter MAP. The converter accesses the data record DATMOD.

The communication unit KOM provides the communication services required for communication with the data processing unit PU2 and with the database manager DBM2.

The data record DATMOD contains meta-data which describes the data models of the two databases DBMS1 and DBMS2.

Using knowledge of the data models of both the databases DBMS1 and DBMS2, the converter MAP converts messages which concern changes to data in one i:"0 data model into messages concerning the change of this data in the other data model.

"The database processes PDB represent processes of the database manager DBM1.

The database manager DBM1 has knowledge about the allocation of synchronisation elements to a given record of the data managed by it. This data record here determines the data which is to made visible to the data processing unit PU2. If now, from one of the processes P1 to P3, a database access INSERT is made "i "which causes the alteration of data, the database processes PDB which are working on this database access check whether a synchronisation element is allocated to this data.

If this is the case, they send a message TRIGGER to the converter MAP and wait oo. before storing the change into the database store DB1 until an acknowledge message ACK is received from the communication unit KOM. If this is not the case, this change 14 of data is not stored in the database store DB1.

The message TRIGGER describes this change of data according to the data model of database DBMS1 and it is converted by the converter MAP into a message STORE which describes the corresponding change to data according to the data model of the database DBMS2. From the message STORE, the communication unit KOM generates corresponding messages which put these changes into effect in the second database, and sends these messages to the database manager DBM2. When this change of data has been entered into the database DBMS2, the database DBMS2 sends corresponding acknowledge messages which are received by the communication unit KOM and sent on to the database processes PDB as an acknowledge message ACK.

The acknowledge message ACK can also be omitted.

It is also possible that the database DBMS2 is a distributed database with many copies of the data to be changed, and that only a particular group of processes respectively has access to one of the copies. With such a database it would be advantageous if the communication unit KOM had knowledge about the distribution of the replicated data. It can then send targeted messages to that portion of the database manager DBM2 which is responsible for the management of the respective database store.

Figure 3b shows the database manager DBM1 and the data processing unit PU2 which exchange messages with each other. The database manager DBM1 has the database processes PDB, the converter MAP, the data record DATMOD and the .communication unit KOM. The communication unit KOM exchanges messages with the data processing unit PU2 and sends messages to the database processes PDB via the converter MAP. The converter accesses the data record DATMOD.

The comments made with respect to Figure 3a apply for the database processes PDB, the converter MAP, the data record DATMOD and the communication unit KOM.

From the data processing unit PU2, the communication unit KOM receives a message which indicates to it the change of data in the database DBMS2. It sends this message on to the converter MAP, which converts this message into a message INSERT which describes the corresponding change of data according to the data model of the database DBMS2. The message INSERT is sent to the database processes PDB and causes these to store the change in the database DBMS1.

The fundctions in the database manager DBM2 are carried out corresponding to Figure 3a and Figure 3b, with the difference that there the converter MAP is omitted.

It is also possible that changes of data in the database DBMS2 are not reported by the database manager DBM2 to the database manager DBM1. Internal messages of the database DBMS2 would then be evaluated by the communication unit KOM and conveyed to database processes if they concerned changes to replicated data.

Through this replication mechanism it is achieved that only the changes of specific data are advised to the other one of the database managers DBM1 and DBM2 and thus that the quantity of data which has to be exchanged for synchronisation is kept small. Furthermore, a high degree of currency for the synchronisation is achieved.

It is also possible that the synchronisation of the databases is not ensured via such a replication mechanism but that, for example, at regular intervals all the data of the databases DBMS1 and DBMS2 needed for cooperation is exchanged. The use of a replication mechanism, however, has the advantage that the data processing units PU1 and PU2 are then coupled via a linkage which is capable of satisfying high speed requirements.

It is also possible not to use the control unit CONTR for the control of the exchange EX, but to use it for any other arbitrary control tasks. Correspondingly, the system platforms and the distribution of tasks of the data processing units PU1 and PU2 would then also be altered.

In a second design example a further possibility for the linking of the databases DBMS1 and DBMS2 will be explained with the aid of Figure 4.

Figure 4 shows the data processing units PU1 and PU2 with respectively the processes P1 to P3, the database DBMS1 and the operating system OS1, and the processes P4 to P6, the database DBMS2 and the operating system 052. The databases DBMS1 and DBMS2 contain respectively the access interface INT], the database manager DBM1 and the database store DB1, and the access interface INT2, *the database manager DBM2 and the database store DB2. The access interfaces INT1 ".00 o and INT2 exchange data with each other.

The data processing units PU1 and PU2 are constructed as in Figure 2. Only 16 the linking of the databases DBMS1 and DBMS2 is implemented by a different mechanism.

The data required for the action on the other data processing unit is not stored in the own database as in the first design example, but is synchronised by a replication mechanism. In this design example, the access to this data takes place through the exchange of messages between the access interfaces INT1 and INT2. When the access interface INT1 receives an access request in the database access language SQL for data from the data processing unit PU2, the access interface INT1 detects this, carries out a conversion of the data model and conveys the request to the access interface INT2. This then requests the corresponding data from the database manager DBM2 and transfers the data to the access interface INT1 as soon as it is received.

With such a linking of the databases DBMS1 and DBMS2, it is advantageous for the databases DBMS1 and DBMS2 to be similar databases and, therefore for the access interfaces INT1 and INT2 to support the same database access language. If this is not the case, in some circumstances a complete conversion between such database access languages is then not possible.

In a third design example an additional possibility for linking the databases DBMS1 and DBMS2 is now explained by means of Figure Figure 5 shows the data processing units PU1 and PU2 respectively with the processes P1 to P3, the database DBMS1 and the operating system OS1, and the processes P4 to P6, the database DBMS2 and the operating system OS2. The databases DBMS1 and DBMS2 contain respedctively the access interface INT1, the database manager DBM1 and the database store DB1, and the access interface INT2, the database manager DBM2 and the database store DB2. Furthermore, both databases are provided with a common database store DB3.

The data processing units PU1 and PU2 are constructed as in Figure 2.

However, contrary to the first design example, the data processing units PU1 and PU2 here have a common hardware platform, i.e. both data processing units PU1 and PU2 are working on the same computer system.

The linking of the databases DBMS1 and DBMS2 is here achieved via the 1.30 common database store DB3. In the database store DB3 all the data is stored which is required for affecting the respective other data processing unit. Since access to this 17 common data processing unit is possible for both database managers DBM1 and DBM2, access becomes possible in this way to data from the respective other data processing unit according to the own data model. For the synchronisation of the access a synchronisation mechanism is necessary here, such as is for example used with operating systems when several processes have access to an operating facility used in common.

flee a ee The claims defining the invention are as follows: 1. A method for linking data processing units which each operate with an independent operating system, wherein in each of the data processing units an independent database is provided which manages at least a part of the data of the respective data processing unit according to a corresponding data model, that a first database of a first data processing unit and a second database of a second data processing unit are linked in such a way that the first database accesses a data record of the second data processing unit which is managed by the second database, and the first database makes this data processable according to its data model for processes of the first data processing unit and that, for cooperation between the first and second data processing units, the processes of the first data processing unit, by accessing the first database, access data from the data record of the second data processing unit.

2. A method as claimed in Claim 1, wherein the data processing units operate with different operating systems and databases which are adapted to the respective tasks.

3. A method as claimed in Claim 1, wherein the data record is accessed by the first and the second database according to different data models and that a data model conversion is carried out for the linking.

4. A method as claimed in Claim 1, wherein the processes of the first data processing unit access semipermanent variables of processes of the second data 20 processing unit.

5. A method as claimed in Claim 1, wherein the first and the second database provide different access procedures for data, and that the processes of the first data processing unit access the data record with the access procedures of the first database.

6. A method as claimed in Claim 1, wherein for the linking, the first and the '5 second database mutually replicate a part of the data managed by them and that the processes of the first data processing unit, via the first database, access a synchronised copy, managed by the first database, of the data record of the second data processing unit.

S7. A method as claimed in Claim 6, wherein a synchronising element is .0 associated with the respective data of the data record in the first database, and that when data which has a synchronising element associated with it is changed, a

Claims

8. A method as claimed in Claim 7, wherein when data with an associated synchronising element is changed, these changes are only entered into a data base store by the database when an acknowledge message has been received from the second database that these changes have been made.
9. A method as claimed in Claim 6, wherein internal messages of the first database are evaluated and a corresponding message is sent to the second database when data of the data record is changed by the first database. A method as claimed in Claim 1, wherein for the linking, the databases access a common data base store in which the data record is stored.
11. A method as claimed in Claim 1, wherein the first database detects an access to the data of the data record and, by means of corresponding access commands, causes the second database to be accessed.
12. A method for controlling an exchange, wherein two or more data processing units working with different operating systems cooperate for the performance of switching functions, that in each of the data processing units an independent database is provided which manages at least a part of the data of the respective data processing unit according to a respective data model, that a first database of a first data processing unit and a second database of a second data processing unit are linked in such a way that the first database accesses a data record of the second data processing unit which is managed by the second database, and makes this data """"visible according to its data model for processes of the first data processing unit and Sthat, for cooperation between the first and second data processing units, the processes of the first data processing unit, by accessing the first database, access data from the data record of the second data processing unit.
13. A first data processing unit with at least one computer system as a hardware platform, with an independent operating system and with a number of control programs at least one of which is so designed that, in its operation, it cooperates with at least one data processing unit which in turn is provided with an independent operating system, wherein the first data processing unit is provided with a first .:30 database for managing the data in accordance with a first data model, that the first database is linked to a second database of the second data processing unit which is so designed that it manages at least a part of the data of the second data processing unit in accordance with a second data model, that the first database is so designed that it accesses a data record of the second data processing unit where the data is managed by the second database and made visible to the processes of the first data processing unit in accordance with the first data model, and that the control program is so designed that during its execution, in order to cooperate with the second data processing unit, it accesses data of the data record of the second data processing unit by accessing the first database.
14. A control unit with at least one computer system as a hardware platform and with at least a first and a second data processing unit, where the first and second data processing unit are each respectively provided with an independent operating system and with a number of control programs of which at least one control program of the first data processing unit is so designed that in its execution it cooperates with the second data processing unit, wherein the first and the second data processing unit are provided with a first and a second database which are respectively so designed that they respectively manage at least a part of the data of the respective data processing unit in accordance with a respective data model, that the first database is linked to the second database, that the first database is so designed that it accesses a data record of the second data processing unit, whose data is managed by the second database, and makes it visible in accordance with its data model for processes of the first data processing unit, and that the control program is so designed that during its execution, for cooperation with the second data processing unit, by accessing the first database, It accesses data of the data record of the second data processing unit. An exchange with a switching network and a control unit, wherein the control unit is a control unit in accordance with Claim 14. :25 16. A method for linking data processing units, substantially as herein described with reference to Figures 1 5 of the accompanying drawings.
17. A method for controlling an exchange, substantially as herein described with reference to Figures 1 5 of the accompanying drawings. 0000 0 0• 1 1 1 21
18. A first date processing unit substantially as herein described with reference to Figures 1 5 of the accompanying drawings.
19. A control unit substantially as herein described with reference to Figures 1 5 of the accompanying drawings. DATED THIS FIFTH DAY OF SEPTEMBER 1996 ALCATEL N.V e *o* 1 I ABSTRACT For the construction of larger control systems, the cooperation is necessary of data processing units (PU1, PU2) which operate with different operating systems (OS1, OS2). For this purpose, the data processing units (PU1, PU2) access each other via a linking of their databases (DBMS1, DBMS2). By means of this linking, data from the other data processing unit (PU1, PU2) is made visible according to the data model of the respective own database (DBMS1, DBMS2). The databases (DBMS1, DBMS2) are here linked via a replication mechanism which functions between their database managers (DBM1, DBM2). This method of linking data processing units (PU1, PU2) is used for the construction of the control unit of an exchange out of different data processing units which cooperate for the provision of switching functions. S