D. C. Oppen and Y. K. Dalal. The Clearinghouse: A decentralized agent for locating named objects in a distributed environment. Xerox Technical Report: OPD-T8103, 1981.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... are in many cases constructed not as part of any particular algorithm but for direct use as the building blocks of various storage and retrieval mechanisms, such as distributed dictionaries, name servers in communication networks, bulletin boards, resource allocation managers and the like [108,95,49,85,116,90,65]. The function common to all of these mechanisms is supplying facilities for storing accumulated information in the system and making it available to potential users throughout the system. The topic of distributed data structures is rather wide, and the distributed setting raises several issues ....

D. Oppen and Y.K. Dalal. The clearinghouse: a decentralized agent for locating named objects in a distributed environment. Technical Report OPDT8103, Xerox Corp., Oct. 1981.

....To enable IFNS to scale well, we designed a defined name space as well as name management and name resolution algorithms, as described in the following section. 4. 2 Name Space Generally, in distributed systems like the Internet, the object locating mechanism is provided by the naming system [12][13] The essential feature in designing the naming system is how to assign a name to an object. This is because the semantics and syntax of the assigned name are directly related to the structure of the name space, and affect the sealability of the name management and resolution algorithms. Once ....

D. Oppen and Y. Dahl, "The Clearinghouse: A Decentralized Agent for Locating Named Objects in a Distributed Environment", Xerox Office Products Division Technical Report, OPD-T8103, 1981.

....Papers on distributed problems are also mixed in with those on concurrency, reliability, and transactions. Bada86 Jaco86 Tane85] Object management: These papers talk about object oriented databases and databases for managing objects. The Oppen paper discusses object naming issues. [Baro81 Derr85 Gold80a Lyng84 Maie86 McLe85 Oppe83 Skar86 Zdon84 Zdon86a] Memory: These are all papers on garbage collection. The Almes paper specifically addresses object oriented systems. There is also some material in the Krasner Smalltalk book. Alme80 Deut76 Dijk78 Lieb81] Message passing: Actors are processes that communicate by passing messages, and specifying ....

D.C. Oppen and Y.K. Dalal, "The Clearinghouse: A Decentralized Agent for Locating Named Objects in a Distributed Environment", ACM TOOIS, vol. 1, no. 3, pp. 230-253, July 1983.

....that we are not tied to a central naming service, which would be inappropriate in an ad hoc network. Given the proximate nature of embedded networking, knowing about things that are not nearby is less important. Some of the ideas about this approach to resource description were first outlined in [12]. Discussion W e have built a prototype embedded network, Piconet. It is one component of a framework of systems and services being envisioned and developed at ORL to support our view of mobility and communications. With Piconet we can demonstrate that radio is the preferred medium for this ....

D. C. Oppen and J. K. Dalal, "The Clearinghouse: A Decentralized Agent for Locating Named Objects in a Distributed Environment," ACM Trans. Office Info. Sys., July 1993. Additional Reading

....and to use this information to optimize network resources. Section 5 presents another protocol that, when combined with gossiping, deals with arbitrary host failures and partitions. 2 Basic Protocol The basic protocol is simple. It is based on gossiping as pioneered in the Clearinghouse project [8, 12]. Long before that, Baker and Shostak [4] describe a gossip protocol, using ladies and telephones in the absence of the widespread availability of computers and networks. In gossip protocols, a member forwards new information to randomly chosen members. Gossip combines much of the e#ciency of ....

Derek Oppen and Yogen Dalal. The Clearinghouse: A decentralized agent for locating named objects in a distributed environment. ACM Transactions on O#ce Information Systems, 1(3):230--253, July 1983. 13

....its old semantics can be applied again during the Two Phase Commit process to tolerate failures if more than the minimal quorum has eventually committed. The ability to ignore several participants can be very useful in large scale distributed systems. A system like The ClearingHouse ([Oppen Dalal 1981]) maintains a naming database over more than a thousand sites, and the likelihood of a failure occurring during the lifetime of a transaction can not be overlooked. The ClearingHouse solved this problem by not using atomic transactions. With the scheme presented here, atomic transactions can be ....

D. C. Oppen and Y. K. Dalal. The Clearinghouse: a decentralized agent for locating named objects in a distributed environment. Technical Report OPD-T8103, Xerox Office Products Division, October 1981. 111

....example, if a key item pair is replicated among n repositories, and if the initial quorum for Lookup consists of a single. node, then the final quorum for Insert must consist of all n repositories. In existing name servers that use replication, such as Grapevine [Birre181] and Clearinghouse lOPpen 81] this trade off is considered to be unacceptable. Instead, a key item pair is added or deleted at a single DM, and the update is subsequently propagated to the other DM s. This approach has the advantage that updates are faster and more likely to succc,d, but it has the disadvantage that the ....

Oppen, D., Dalai, Y, K., "The clearinghouse: a decentralized agent for locating named objects in a distributed environment," Xerox technical report OPD-T8103, October, 1981.

....as the number of names increased instead of adding to the existing ones. Increasing the size of the registries would have required larger servers, which were quite small by today s standards. Creating more increased the possibility that an individual would move between registries. Clearinghouse [17], the commercial successor to Grapevine, had 22 better scaling characteristics at the expense of adding a level to the naming hierarchy. This creeping partitioning of the namespace reduces the longevity of the names. 2.2 Lampson s Global Name Service Lampson used Grapevine and Clearinghouse as ....

D. C. Oppen, Y. K. Dalal, The Clearinghouse: A Decentralized Agent for Locating Named Objexts in a Distributed Environment, Xerox OPD-T8103, October 1981.

....systems protocols compare to Bayoffs based on the sketchy information available. Grapevine, one of the earliest weakly replicated systems, propagated updates via electronic mall [2] Electronic mall is not completely reliable, however, so the product version of Grapevine, called Clearinghouse [15], added a background antientropy process in addition to mall delivery. It was later realized that epidemic style algorithms, like Clearinghouse s anti entropy, could be used by themselves to fully propagate updates [3] Pairwise reconciliation of replicas is currently used in several systems ....

....exchange updates. However, update propagation is more affected by communication outages. In many systems using lazy replication, the information exchanged between replicas is based on data objects with associated update timestamps or version vectors. This is true for Grapevine [2] Clearinghouse [15], Notes [10] and Microsoft Access [8] File systems like Coda [18] and Ficus [7] exchange updated files between servers or between clients and servers. The notion of reconciling logs of update operations held at various replicas, as is done in Bayou via the anti entropy protocol, has been ....

[Article contains additional citation context not shown here]

D.C. Oppen and Y. K. Dalai. The Clearinghouse: A decentralized agent for locating named objects in a distributed environment. ACM Transactions on Office Information Systems 10):230-253, July 1983.

....to intra region latency is high in the generated topologies. Consequently, a repair is sent in regional multicast only if the entire region missed the message. VI. RELATED WORK Randomization was previously used in epidemic algorithms to disseminate updates in a distributed database environment [22][16] More recently, van Renesse et al. proposed a failure detection service using the random gossiping technique [23] The error recovery algorithm in RRMP combines our previous work on randomized error recovery in the Bimodal Mul 3 4 5 6 7 8 9 10 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 ....

Derek C. Oppen and Yogen K. Dalal, "The Clearinghouse: A decentralized agent for locating named objects in a distributed environment," Tech. Rep., Xerox, 1981.

....attributes may be expensive and difficult. At present it is only implemented on VMS using the DNA protocols, but will be moved to UNIX and TCP in the next year. Sollins [Page 9] RFC1107 RFC.net Page 10 of 19 RFC 1107 A Plan for Internet Directory Services July 1989 Clearinghouse This service [7,2] is part of the Xerox network environment. It operates today as a global service for Xerox. They have considerable experience with its operation, including problems of scale. Clearinghouse provides a three level hierarchy of names that are mapped to sets of properties. Loose consistency is ....

Oppen, D., and Y. Dalal, "The Clearinghouse: A Decentralized Agent for Locating Named Objects in a Distributed Environment", Tech. Rept. OPD-T8103, Xerox Corporation, Palo Alto, CA, October 1981.

....Its namespace was divided into registries of names with each registry being stored on one or more registration servers. This hierarchical naming scheme does not match our need for a flat space, nor does the replication of entire registries scale to global proportions. The Clearinghouse system[6] increased the scalability of Grapevine by adding a third level to the naming hierarchy. The names were broken down by organization and then domain. The distribution of the servers followed the structure of the namespace such that each domain required a clearinghouse server. The Domain Name ....

Derek C. Oppen and Yogen K. Dalal. The Clearinghouse: A Decentralized Agent for Locating Named Objects in a Distributed Environment, Xerox Corporation, October 1981.

....the structures, the operations and the semantics. Implementation issues will be given considerations in section 5. Section 6 presents a performance study of the filtering mechanisms, and finally, we conclude with discussions on future directions in Section 7. 2 Data Replication In the literature [3, 11, 11, 4, 5, 10, 12, 14, 17, 16], data replication can improve data availability and system performance in distributed file systems. In such systems, the information about the location and the content of a replica is available to all participating sites. Data in replicas will be shared by all the participating sites and will be ....

D.C. Oppen and Y.K. Dalal. The Clearinghouse: A decentralized agent for locating named objects in a distributed environment. ACM Transactions on Office Information Systems 1(3):230-253, July, 1983.

....levels and uses different replication algorithms. Currently, about 350 Clearinghouse services are in operation all over the world. The main problem resulting from this size is the long propagation of updates (days or weeks) and the heavy Clearinghouse induced internetwork traffic. Reference: [347] CMCFS CMCFS (Carnegie Mellon Central File System) was designed in early 1980. It provides two kinds of update schemes. First, the immutable file approach by which a new version is created with each update and, second, a traditional update based on transactions. CMCFS is part of the Spice ....

D.C. Oppen and Y.K. Dalal, "The Clearinghouse: A Decentralized Agent for Locating Named Objects in a Distributed Environment", ACM Transactions on Office Information Systems, 1(3):230--253, July 1983.

....moderate server loads, the problem was finding the desired network resource knowing only its name or property. Many techniques were investigated, including: 1) the use of multicast or broadcast communication to touch all the locations where the resource may reside in an attempt to find it (e.g. [4, 33]) 2) the use of various name server architectures to lookup the location of the resource (e.g. 31, 21, 6] and 3) the use of caching a resource s location (not content) at sites where the resource is frequently accessed [41] This early work typically dealt with a single instance of the ....

D. Oppen and Y. Dalal. The clearinghouse: A decentralized agent for locating named objects in a distributed environment. ACM Transactions on Office Information Systems, 3(1):230--253, July 1983.

....moderate server loads, the problem was finding the desired network resource knowing only its name or property. Many techniques were investigated, including: 1) the use of multicast or broadcast communication to touch all the locations where the resource may reside in an attempt to find it (e.g. [3, 24]) 2) the use of various name server architectures to lookup the location of the resource (e.g. 22, 15, 4] and 3) the use of caching a resource s location (not content) at sites where the resource is frequently accessed [32] This early work typically dealt with a single instance of the ....

D. Oppen and Y. Dalal. The clearinghouse: A decentralized agent for locating named objects in a distributed environment. ACM Transactions on Office Information Systems, 3(1):230--253, July 1983.

....known to a portion of a distributed system is propagated to the rest of the system. Diffusion is useful for driving replicated data toward a consistent state over time, and has found application for this purpose, e.g. in USENET News [LOM94] and in the Grapevine [BLNS82] and Clearinghouse [OD81] systems. The quality of a diffusion protocol is typically defined by the delay until the update has This paper appears in Proceedings of the 18th IEEE Symposium on Reliable Distributed Systems (SRDS 99) October 1999, EPFL, Lausanne, Switzerland, pages 134 143. y AT T Labs Research, ....

D. C. Oppen and Y. K. Dalal. The Clearinghouse: A decentralized agent for locating named objects in a distributed environment. Xerox Technical Report: OPD-T8103, 1981.

....to intra region latency is high in the generated topologies. Consequently, a repair is sent in regional multicast only if the entire region missed the message. VI. RELATED WORK Randomization was previously used in epidemic algorithms to disseminate updates in a distributed database environment [20][14] More recently, van Renesse et al. proposed a failure detection service using the random gossiping technique [21] The error recovery algorithm in RRMP protocol combines our previous work on randomized error recovery in the Bimodal Multicast protocol [9] and hierarchical error recovery ....

Derek C. Oppen and Yogen K. Dalal, "The Clearinghouse: A decentralized agent for locating named objects in a distributed environment," Tech. Rep., Xerox, 1981.

....only to objects, but to contexts as well, thus providing multiple levels of indirect reference. This last concept is in fact the focus of the majority of subsequent work, as it introduces many potential problems to the resolution process. The notion of naming indirection was first formalized in [50], which models relationships between objects, names, and contexts as directed graphs. Each node in such a graph represents an object or a context. Each edge is labeled with exactly one name. The set of bindings that make up a context are thus represented by the set of outgoing edges of the ....

....of names. We have omitted such properties because we have not found them to be significant in determining the patterns of communication that naming systems are capable of expressing. However, it is possible to augment the taxonomy with an orthogonal set of resolution properties. Previous work[19, 50] has already produced very useful analytical models of partitioned, hierarchical name spaces, which are the source of most resolution complexity. This work would serve as a starting point for determining resolution properties. One way to view the naming system properties we have defined is as a ....

D. Oppen and Y. Dalal, The clearinghouse: a decentralized agent for locating named objects in a distributed environment, XEROX Office Products Division Research Report, Oct. 1981

....1 components. Moreover, a naming system must provide trustworthy service even though parts of the system may belong to many autonomous and mutually suspicious groups different departments, corporations, and even countries. Some progress on this large scale design problem has been reported [2, 13, 16, 19, 22]; however, the performance of these systems appears inadequate for file naming at best, it is adequate for naming hosts, mailboxes, and other relatively infrequently accessed objects. Name lookup operations are a significant factor in system performance. For instance, Leffler et al. 17] ....

....levels of our naming design, under the assumption that directories at the global level can be implemented using known name server technology. Perhaps the most advanced work of that kind to date has been Lampson s design [1, 16] an outgrowth of the work on Grapevine [2, 27] and the Clearinghouse [22]. The Domain Name service [19, 20] of the DARPA Internet is a more limited system in the same class. 21 Additional work in this area is surveyed in Terry s thesis [30] These global name services are typically used to map from names to unique identifiers for hosts, mailboxes, or services. As ....

D. C. Oppen and Y. K. Dalal. The Clearinghouse: A decentralized agent for locating named objects in a distributed environment. ACM Transactions on Office Information Systems, 1(3):230--253, July 1983.

No context found.

D. C. Oppen and Y. K. Dalal. The Clearinghouse: A decentralized agent for locating named objects in a distributed environment. Xerox Technical Report: OPD-T8103, 1981.

No context found.

D.C. Oppen and Y.K. Dalal. The clearinghouse: a decentralized agent for locating named objects in a distributed environment. ACM Transactions on O#ce Information Systems, 1(3):230--253, 1983.

No context found.

Oppen, D. C., and Dalal, Y. K. The Clearinghouse: A decentralized agent for locating named objects in a distributed environment. ACM Transactions on Office Information Systems. 1(3), July 1983.

No context found.

Oppen, D. and Y. Dalal. The clearinghouse: A decentralized agent for locating named objects in a distributed environment. Technical Report Tech. Rep. OPD-78103, Xerox Office Products Division, 1981.

No context found.

D. C. Oppen and Y. K. Dalal. The Clearinghouse: A decentralized agent for locating named objects in a distributed environment. Xerox Technical Report: OPD-T8103, 1981.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC