Distributed data processing system

• US 4,430,699 A
• Filed: 02/17/1981
• Issued: 02/07/1984
• Est. Priority Date: 02/15/1980
• Status: Expired due to Term

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: US Philips Corporation (Koninklijke Philips N.V.)
• Inventors: Segarra, Gerard, Phulpin, Francois J.
• Primary Examiner(s): Nusbaum, Mark E.
• Assistant Examiner(s): Harkcom, Gary V.
• Application Number:
  US06/235,291
• Time in Patent Office: 1,085 Days
• Field of Search: 364/200 MS File, 364/900 MS File, 350/96.15, 350/96.16
• US Class Current: 709/230
• CPC Class Codes: G06F 15/17337 Direct connection machines, e.g. completely connected computers, point to point communication networks coupling between buses G06F13/4004
• CPC Class Codes: H04L 12/43 with synchronous transmission, e.g. time division multiplex [TDM], slotted rings
• CPC Class Codes: H04L 12/56 Packet switching systems

Abstract

A distributed data processing system having several local systems (LS) between which communication is provided via clearly defined functional layers. A functional coordination layer managed by systems intercommunication processors (SIP) is responsible for the functions of coordination, communication, control, initialization and simulation relative to the LSs. A functional communication layer managed by communication modules (CM) is responsible for the communication protocols between the LSs. A functional transport layer including transmission modules (TM), a looped optical bus and a looping unit (LIG) element. A description of each LS is contained in its SIP in the form of descriptive tables facilitating the use of the resources of the global system by any LS via parameter translations, local to global to local, each communication between LSs being performed in the form of a transaction having separate interrogation, auto-selection, presentation, processing and result return phases.

Claims

1. A distributed data processing system comprising a general communications network and a plurality of local systems, wherein the network is passive with respect to the self-contained operation of the respective local systems, and wherein each local system comprises at least one central processing unit, associated memory, at least one peripheral device and at least one residential process, local monitor means and a monitor extension module (42), wherein the control of the data processing in a local system is distributively effected by the local system itself, wherein each local system is interfaced to the network via its monitor extension module by means of a systems intercommunication processing means, the set of systems intercommunication processing means constituting a distributed coordination, communication, control, initialization and simulation means for the data processing system as a whole, each system intercommunication processing means being attached to the network via a communication module, the set of communication modules constituting a distributed control means for controlling communication protocols on the network for the data processing system as a whole, each communication module having sequencing means for controlling, in the communication protocols in a first sequence of steps, an addressed logical link from a source local system to a single destination local system and controlling in a second sequence of steps a broadcast logical link from a source local system to one or more destination local systems, said first and second sequences comprising conditionally executable logical link establishing steps, logical link maintaining steps and logical link terminating steps, said communication modules furthermore comprising information flow rate controlling means, presenting means for presenting the same general order of selected events at each local system, and error control means for detecting and thereupon recovering data communication errors, each communication module being plugged into the physical network by means of a transmission module, the set of transmission modules in combination with the physical network and a looping unit constituting a physical transport layer means, the physical network being a looped optical bus which is looped by said looping unit, wherein each transmission module comprises first control means for controlling parity errors, second control means for controlling synchronization between the transmission module itself and the transfer on the optical bus as presented, and bidirectional conversion means between electrical and optical signals, and wherein the looping unit has a first converting means for converting optical signals received from the bus to electrical signals, processing means for processing said electrical signals, and second converting means for converting processed electrical signals to optical signals for insertion into said optical bus, said processing means having encoding/decoding means for each transmission line of said bus, initializing means for initializing the transmission synchronization on the optical bus, and frame control means for controlling transmission and reception of data in a frame of a cycle of frames, the frame in question being allocated to one local system.

2. A system as claimed in claim 1, in which said transmission module (TM 15) also comprises:

(a) a circuit for calculating the parity (169) connected to a data transmission coder (170) which in turn is connected via a shift register (RDS

     171), a delay component (τ

     172) and an analog circuit (CA

     173) to a laser diode (DL

     174), said DL, controlled by an element (POLDL

     175), transmitting the data after electro-optical conversion to said optical bus (16),

(b) an avalanche photodiode (PDA

     176) established by a converter (ALHT

     177) to perform the opto-electronic conversion of the data from said optical bus (16), said converted data being amplified in an amplification circuit (DAE

     178) and then processed non-linearly in a circuit (TNL

     181), the output of said TNL (181) being applied to a phase locking loop (PLL

     182), the outputs of said PLL (182) being applied in turn to a decision circuit (183), a shift register (RDE

     184) controlled by said decision circuit (183) performing the series-parallel transformation of said reception data, and said RDS (171), the output of said RDE (184) being linked to a decoder (185) and a comparator circuit (186),

(c) a state sequencer (SEQ

     187) managing the emission and reception frames in accordance with the information received from said PLL (182), decoder (185) and CM (13),

(d) a command integrator COM TRIAC (179) connected to the output of said DAE (178) and controlling a triac (180), said triac (180) comprising means for powering said CM (13) and TM (15).

3. A system as claimed in claim 1, in which said looping unit (LIG 17) also comprises:

(a) a transmission shift register (RDE

     191) connected to an application circuit (CA

     195), the output of said CA (195) being connected to a laser diode (DL

     196) controlled by a regulating element (REG

     197), said DL (196) transmitting data in optical form to said optical bus (16), said RDE (191) being loaded either by a FIFO register (193) or by bits issued by a PROM (192),

(b) an avalanche photodiode (PDA

     199) controlled by a converter (ALHT 197 a) to receive the data from said optical bus (16), the output of said PDA (199) being regenerated by a regeneration circuit (DAE 199a), said circuit DAE (199a) being in turn connected to a decision circuit (198) making it possible to load said FIFO register (193) via a reception shift register (RDR 193a),

(c) a master clock (H

     190) emitting pulses at bit frequency to synchronize said RDE (191), RDR (193a) decision circuit (198) and a state sequencer (SEQ

     194), said SEQ (194) ensuring said initialization and transmission synchronization procedures and controlling said RDE (191), PROM (192), FIFO register (193) and RDR (193a).

4. A system as claimed in claim 3 in which a description of the resources available at each local system (LS 10) is loaded into the RAM (54) of each of said intercommunication processor (SIP 11) when said global distributed system is generated, each of said descriptions of the resources available at each LS being dynamically updated in relation to the needs of the users and the evolution of said distributed system.

5. A system as claimed in claim 4, in which each of said descriptions of the resources available at each LS also comprises parameter translation and source code correspondence tables making it possible to translate between said LSs, a descriptive table of system resources and domains of application, a descriptive table of the communication mail-boxes, descriptive tables of the user resources, a table of free local codes, and chain start address tables of global and file code numbers.

6. A system as claimed in claim 5, in which the processing of each service request from a user process located in an LS is performed by means of a transaction also comprising discrete phases of interrogation, auto-selection, presentation, processing and return of the result, said service request sent being reformated by said monitor extension module (EM 42) and communicated to the originating SIP (11) across the LS/SIP interface by a block of commands and said LS/SIP intercommunication mechanisms.

7. A system as claimed in claim 6, in which said interrogation phase also comprises means for said originating SIP (11) to communicate an interrogation in the addressed mode to the LS (10) concerned when there is no possible choice and when the user process can localize the requested resource, means for said originating SIP (11) to communicate said interrogation in broadcast mode to all the LS when there is a possible choice of M among N resources, where N is greater than M, or when the user process cannot localize the resource, said service request being formatted as an interrogation block by said originating SIP and communication either in addressed or broadcast mode to the addressed LS or to all the LSs through said functional coordination (12) communication (14) and transport layers (18) using said intercommunication mechanism between said layers.

8. A system as claimed in claims 6 or 7, in which the auto-selection phase also comprises means, when there is a choice, of M among N resources, for selecting M LSs, means for each LS to analyze the replies sent to determine which M LSs are selected, means for the (N-M) LS not selected to cancel said service request, said auto-selection mechanism permitting all the LSs to have the same view of the global order of events.

9. A system as claimed in anyone of claims 6 or 7, in which said presentation phase also comprises means for presenting to the selected LS a service request in the form of a block of commands, said block of commands being communicated by the originating SIP to the selected LS through said coordination, communication and transport layers (18) using said interconnection mechanism between said layers, said EM located in each selected LS simulating said service request by generating the image of a local process.

10. A system as claimed in anyone of claims 6 or 7, in which said processing phase also comprises means allowing the local monitor of each selected LS to process said request in the same way as when it is issued by a local process, together with interruption means to inform said EM of the end of the processing of said request.

11. A system as claimed in anyone of claims 6 or 7, in which said result return phase also comprises means allowing the selected LS to communicate the result of the processing of said request to its SIP via the LS/SIP interface, means allowing said SIP to construct a result block as a function of the processing performed, said result block being communicated to the originating SIP via said coordination, communication and transport layers, said originating SIP, after the activation of said EM, transferring said result to the local user process via the SIP/LS interface.