J. Ordille, B. Miller. "Nomenclator Descriptive Query Optimization in Large X.500 Environments," ACM SIGCOMM Symposium on Communications Architectures and Protocols, pp.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....appropriate extensions, the DNS could be used to support many of the applications mentioned above. However, it has been decided that a higher level of directory service is needed. Currently, several directory service pilots based on X. 500 are being developed and evaluated in the Internet (e.g. see [20, 19, 10]) For convenience, we refer to the X.500 directory as the Directory throughout the rest of the paper. The Directory information database is organized using a hierarchical, tree structured model referred to as the Directory Information Tree, or DIT. In the DIT, each entry, or node, represents ....

....than is strictly necessary. They also show that negative caching, i.e. caching bad names or the absence of resource records, does not improve performance. Nomenclator, a system which uses X. 500 as its data repository to provide efficient attribute based naming in an internet, is described in [20]. Nomenclator increases query performance up to 40 times that of the Directory by trimming branches of the DIT and by caching results from previous queries as well as characteristics of these queries. Techniques for managing caches of hints, i.e. caches of data that are not completely accurate, ....

J. Ordille, B. Miller, "Nomenclator descriptive query optimization for large X.500 environments", Proc. ACM SIGCOMM '91, pp. 185-196, Zurich, Switzerland, Sept. 1991.

....Distributed X.500 Table 1: A Taxonomy for Internet Resource Discovery Services. customized views of the file system hierarchy called virtual directories, which help locate and organize information. A semantic file system research prototype has been implemented on top of Sun NFS. Nomenclator [21, 23] implements an attribute based, or yellow page naming on top of hierarchical naming systems. Nomenclator access functions are, in essence, servers that periodically traverse the appropriate portions of the underlying name space and other access functions. They build indices of the objects ....

Joan J. Ordille and Barton P. Miller. Nomenclator descriptive query optimization for large X.500 environments. ACM SIGCOMM 91 Conference, pages 185--196, September 1991.

....combinations of the desired attribute value pairs. Transducers and queries produce customized views of the file system hierarchy called virtual directories, which help locate and organize information. A semantic file system research prototype has been implemented on top of Sun NFS. Nomenclator [9] implements attribute based naming on top of other naming systems. Nomenclator access functions are, in essence, servers that periodically traverse the appropriate portions of the underlying name space and other access functions. They build indices of the objects encountered that satisfy certain ....

Joann J. Ordille and Barton P. Miller. Nomenclator descriptive query optimization for large X.500 environments. ACM SIGCOMM 91, pages 185--196, September 1991.

....in Computing Systems: Volume 5, Number 4, 1992. ALso USC TR 92 519) Wide Web [3] and Prospero [11] organize information into a distributed hypertext. With these tools, people create links between relevant information which may reside on different servers. Tools like archie [7] Nomenclator [14], and netfind [15] build indices from information scavenged from various repositories and sources. This paper describes a discovery architecture that, similar to this second set of tools, automatically clusters pointers to related information obtained from other repositories and discovery services ....

....combinations of the desired attribute value pairs. Transducers and queries produce customized views of the file system hierarchy called virtual directories, which help locate and organize information. A semantic file system research prototype has been implemented on top of Sun NFS. Nomenclator [14] implements attribute based or yellow page naming on top of hierarchical, white page naming systems. A Nomenclator access function is, in essence, a server that periodically traverses other access functions and appropriate parts of the white page name space, and constructs an index of the objects ....

Joan J. Ordille and Barton P. Miller. Nomenclator descriptive query optimization for large X.500 environments. ACM SIGCOMM 91 Conference, pages 185--196, September 1991.

....function, SQL and its familiar Select From Where construct is highly flexible. Especially important is its ability to join information from multiple relations in a database, thus promoting querying of the relationships that exist between the objects being modelled. AID [1] and Nomenclator [10] are two previous attempts to integrate X.500 with relational query languages. AID implements an SQL like interface to QUIPU. Its queries have the Select From Where form, but the From clause may only specify a single table. Each node in a DIT is considered to be a table; the tuples belonging to a ....

J. J. Ordille and B. P. Miller. Nomenclator Descriptive Query Optimization for Large X.500 environments. In Proceedings of the 1991 SIGCOMM Conference, 1991.

No context found.

J. Ordille, B. Miller. "Nomenclator Descriptive Query Optimization in Large X.500 Environments," ACM SIGCOMM Symposium on Communications Architectures and Protocols, pp.

....name service, called Nomenclator. Nomenclator answers queries about people by retrieving information from a variety of name services, including CCSO name services [3] and relational database services. A previous version of Nomenclator improved the performance of descriptive queries in X. 500 [7]. The Nomenclator effort is a continuing research program to develop techniques that improve query speed, preserve organization autonomy, protect privacy and enhance data availability in large scale information systems [7] 6] 5] 4] This paper explores aspects of Nomenclator that relate to ....

....of Nomenclator improved the performance of descriptive queries in X. 500 [7] The Nomenclator effort is a continuing research program to develop techniques that improve query speed, preserve organization autonomy, protect privacy and enhance data availability in large scale information systems [7], 6] 5] 4] This paper explores aspects of Nomenclator that relate to distributed, heterogeneous information gathering. Section 2 identifies two information gathering principles that are central to the active catalog. Section 3 briefly describes techniques for meta data gathering, ....

Ordille, J.J. and Miller, B.P. Nomenclator Descriptive Query Optimization in Large X.500 Environments. ACM SIGCOMM Symposium on Communications Architectures and Protocols (September, 1991), 185-196. ` `

....actual responses, resulting in search times of a few seconds. Even a request to search the global name space for a person with a popular name can be answered in seconds. Distributed active catalogs and meta data caching are currently used in a prototype descriptive name service called Nomenclator [10]. Nomenclator answers selection To Appear 1 ICDCS and projection queries on relations that span heterogeneous name services in the global internetwork. Like the Domain Name System [8] Nomenclator currently uses timestamps to identify and replace potentially stale data and meta data in its ....

....we must analyze the scaling behavior of our framework for query workloads from multiple users. How well will our query processing scale to millions of users Our previous work shows that it is possible to constrain queries in a real environment, the X. 500 name To Appear 6 ICDCS space [10]. In that study, for example, the active catalog constrains the attribute surname with value Miller to only 32 of the X.500 data repositories in the United States. Surname is an attribute users are likely to know when searching for information about people. Moreover, even a common surname like ....

J. J. Ordille and B. P. Miller, "Nomenclator Descriptive Query Optimization in Large X.500 Environments," ACM SIGCOMM Symposium on Communications Architectures and Protocols, Zurich, pp. 185-196 (September, 1991).

....relational flavor like the CSNET name service [11] and sources of public files like Archie [7] into one descriptive (or relational) name service to make resource location simpler for everyone. Several efforts already support descriptive queries over file trees and other hierarchical name spaces [9, 17, 21]. The challenge is not the integration of the data models of the component systems, but the performance of the resulting name service. The most frequent query for any descriptive name service is the global selection query. To answer a global selection query, the name service must locate objects ....

....of local name services, or require excessive software modifications and support. Autonomy exacerbates problems of scale. Scale and performance improve if we create a global index that accepts a selection predicate and constrains it to a subset of the partitions in the global multidatabase [15, 17]. Autonomy makes such an index virtually impossible, because maintaining the traditional level of consistency (serializability) of the index introduces a global synchronization bottleneck and requires changes in the component systems. Interestingly, this tension between autonomy and consistency is ....

[Article contains additional citation context not shown here]

J. J. Ordille and B. P. Miller, "Nomenclator Descriptive Query Optimization in Large X.500 Environments," ACM SIGCOMM Symposium on Communications Architectures and Protocols, Zurich, pp. 185-196 (September, 1991).

....that a name space partition is available is .999, then the probability that 10,000 independent partitions are available simultaneously is .0000452 (or 24 minutes of availability per year) Only 45 of X. 500 partitions in the United States were available at one time during experiments in 1991 [76]. Measurements of the DNS indicate that more than 10 of the partitions at educational institutions are not replicated, so the unavailability of a single system can make one of those partitions unavailable [26] The reasons for unavailability are failure, abandonment, and intermittent ....

.... (e.g. in Archie) or partitioned massive indices (e.g. in WAIS or distributed indices) as the current techniques for supporting large scale descriptive services on the Internet [10] Our previously published work and the work in this thesis provides a middle ground between these two techniques [76, 77]. We provide massive replication of parts of indices through meta data caching at the user s site. Indices from different organizations can be integrated into one (perhaps massive) index by our referral graph techniques, but only parts of this index are replicated in meta data caches. Hierarchical ....

[Article contains additional citation context not shown here]

Ordille, J. J. and B. P. Miller, "Nomenclator Descriptive Query Optimization in Large X.500 Environments, " ACM SIGCOMM Symposium on Communications Architectures and Protocols, Zurich, pp. 185196 (September, 1991).

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