Delta-4: Definition and Design of an open Dependable Distributed Architecture

From Navigators

(Difference between revisions)
Jump to: navigation, search
 
(6 intermediate revisions not shown)
Line 10: Line 10:
|Duration months=69
|Duration months=69
|Keywords=Delta-4
|Keywords=Delta-4
-
|Summary=The Delta-4 project is concerned with the achievement of dependability in open distributed systems, including real-time systems. The project has developed a distributed fault-tolerant architecture for the support of dependable applications for computer integrated manufacturing, process control, office systems, etc.
+
|Summary===Aims==
-
==Aims==
+
The aim of the Delta-4 project was to provide a computational and communication infrastructure for application domains that require distributed system solutions with various dependability and real-time constraints. The scale of distribution in the targeted application domains is commensurate with the distances that can be covered by local area networks.
-
The aim of the Delta-4 project is to provide a computational and communication infrastructure for application domains that require distributed system solutions with various dependability and real-time constraints. The scale of distribution in the targetted application domains is commensurate with the distances that can be covered by local area networks.
+
==Achievements==
-
==Approach and Methods==
+
The Delta-4 project, which ran for 5 years (1986-91) with around 40 Mio Euro funding, is considered one of the most relevant European projects to date, having launched seminal research in fault tolerance in open distributed computing, using modular COTS systems, a pioneering achievement that left an indelible mark in the way F/T systems are designed today.
-
To be able to satisfy a large range of application requirements in a cost-effective manner, the Delta-4 architecture can provide various degrees of dependability and performance. Fault-tolerance is based on the replication of components of run-time software. Replicas are allocated to distinct nodes of a local area network.
+
A seminal publication in the fault tolerance flagship conference FTCS (nowadays DSN):
-
The Delta-4 architecture offers two variants (both based on sub-systems presenting a high degree of commonality):
+
* D. Powell, D. Seaton, G. Bonn, Paulo Verissimo, F. Waeselynck, "[[Publication:D-powell1988the-delta-4-168|The Delta-4 Approach to Dependability in Open Distributed Computing Systems]]", in Proceedings of the 18th IEEE International Symposium on Fault-Tolerant Computing (FTCS), pp. 246-251, June 1988., Jun. 1988.
-
(a) the Delta-4 Open System Architecture (D4-OSA) which, as its name suggests, is an open architecture able to accommodate heterogeneity,
+
has introduced the main features of DELTA-4: the innovative architecture, dependability model and implementation approaches. The paper introduced many concepts and laid down fundamental advancements in the theory and practice of fault-tolerant computing, which are now commonly used in many current dependable systems. It inspired further seminal research inside the project consortium itself, as well as inspiring many researchers and practitioners.
-
(b) the Delta-4 Extra Performance Architecture (D4-XPA) which provides explicit support for assuring timeliness.
+
A brief summary of the s.o.t.a. advances reported in the paper can be found in [[Media:Delta4-project-PressRelease.pdf| the Delta-4 press release]].
-
==Progress and Results==
+
A full account of the results of Delta-4 was cast in book in 1991, and is made available below, provided the users respect the copyright notice in http://www.navigators.di.fc.ul.pt/wiki/Publications.
-
The project has developed and demonstrated prototypes of the architecture at yearly intervals throughout the project. Two pilot demonstrators were implemented: a payment card authorization centre (Credit Agricole) and a computer-integrated manufacturing system for the Renault automobile company. The project ended in December 1991.
+
* [[Media:Chapter 1.pdf|Chapter 1]]
 +
* [[Media:Chapter 2.pdf|Chapter 2]]
 +
* [[Media:Chapter 3.pdf|Chapter 3]]
 +
* [[Media:Chapter 4.pdf|Chapter 4]]
 +
* [[Media:Chapter 5.pdf|Chapter 5]]
 +
* [[Media:Chapter 6.pdf|Chapter 6]]
 +
* [[Media:Chapter 7.pdf|Chapter 7]]
 +
* [[Media:C8.pdf|Chapter 8]]
 +
* [[Media:Chapter_9.pdf|Chapter 9]]
 +
* [[Media:Chapter 10.pdf|Chapter 10]]
 +
* [[Media:Chapter 11.pdf|Chapter 11]]
 +
* [[Media:Chapter 12.pdf|Chapter 12]]
 +
* [[Media:Chapter 13.pdf|Chapter 13]]
 +
* [[Media:Chapter 14.pdf|Chapter 14]]
 +
* [[Media:Chapter 15.pdf|Chapter 15]]
|NavigatorsSite=INESC
|NavigatorsSite=INESC
|Team Size=6
|Team Size=6
-
|Researchers=Paulo Verissimo, Luís Rodrigues, José Rufino,  
+
|Researchers=Paulo Verissimo, Luís Rodrigues, José Rufino,
}}
}}

Latest revision as of 19:14, 19 January 2019


Aims

The aim of the Delta-4 project was to provide a computational and communication infrastructure for application domains that require distributed system solutions with various dependability and real-time constraints. The scale of distribution in the targeted application domains is commensurate with the distances that can be covered by local area networks.

Achievements

The Delta-4 project, which ran for 5 years (1986-91) with around 40 Mio Euro funding, is considered one of the most relevant European projects to date, having launched seminal research in fault tolerance in open distributed computing, using modular COTS systems, a pioneering achievement that left an indelible mark in the way F/T systems are designed today.

A seminal publication in the fault tolerance flagship conference FTCS (nowadays DSN):

has introduced the main features of DELTA-4: the innovative architecture, dependability model and implementation approaches. The paper introduced many concepts and laid down fundamental advancements in the theory and practice of fault-tolerant computing, which are now commonly used in many current dependable systems. It inspired further seminal research inside the project consortium itself, as well as inspiring many researchers and practitioners.

A brief summary of the s.o.t.a. advances reported in the paper can be found in the Delta-4 press release.

A full account of the results of Delta-4 was cast in book in 1991, and is made available below, provided the users respect the copyright notice in http://www.navigators.di.fc.ul.pt/wiki/Publications.

Publications

  • Paulo Verissimo, Luís Rodrigues, José Rufino, “The Atomic Multicast protocol (AMp)”, in DELTA-4 - A Generic Architecture for Dependable Distributed Computing, David Powell, Eds., ser. ESPRIT Research Reports. Springer Verlag, Nov. 1991, ch. 10, pp. 267–294.

  • P. G. Bond, D. Seaton, Paulo Verissimo, “Real-time Concepts”, in DELTA-4 - A Generic Architecture for Dependable Distributed Computing, David Powell, Eds., ser. ESPRIT Research Reports. Springer Verlag, Nov. 1991, ch. 5, pp. 89–124.

  • David Powell, Paulo Verissimo, “Distributed fault-tolerance”, in DELTA-4 - A Generic Architecture for Dependable Distributed Computing, David Powell, Eds., ser. ESPRIT Research Reports. Springer Verlag, Nov. 1991, ch. 6.

  • Paulo Verissimo, P. A. Barrett, P. G. Bond, A. M. Hilborne, Luís Rodrigues, D. Seaton, “The Extra Performance Architecture (XPA)”, in DELTA-4 - A Generic Architecture for Dependable Distributed Computing, David Powell, Eds., ser. ESPRIT Research Reports. Springer Verlag, Nov. 1991, ch. 9, pp. 211–266.

  • Paulo Verissimo, Luís Rodrigues, “Reliable multicasting in high-speed lans”, in High-Capacity Local and Metropolitan Area Networks, ser. NATO ASI. Springer Verlag, Jan. 1991, vol. F72, pp. 397–412.

  • P. A. Barrett, P. G. Bond, A. M. Hilborne, Luís Rodrigues, D. Seaton, N. A. Speirs, Paulo Verissimo, “The Delta-4 Extra Performance Architecture (XPA)”, in From the Digest of Papers, The 20th International Symposium on Fault-Tolerant Computing (FTCS), pp. 481-488, June 1990., Jun. 1990.

BibTeX

Navigators - Delta-4 project
Personal tools
Navigators toolbox