Internet-Draft H. Alvestrand draft-alvestrand-directory-defs-00.txt EDB Maxware Target Category: Informational October 1999 Expires: April 2000 Definitions for talking about directories Status of this Memo The file name of this memo is draft-alvestrand-directory-defs- 00.txt This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet- Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet- Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. Abstract When discussing systems for making information accessible through the Internet in standardized ways, it may be useful if the people discussing have a common understanding of the terms they use. This document is not intended to be either comprehensive or definitive, but is intended to give some aid in mutual comprehension when discussing information access methods to be incorporated into Internet Standards-Track documents. 1. Introduction and basic terms We suggest using the following terms for the remainder of this document: - Information: Something for which one can imagine multiple worlds where the item in question has different values. The fact of which particular value is true for this world is information. Definitions for talking about directories Harald Alvestrand draft-alvestrand-directory-defs-00.txt Expires April 2000 This definition is extremely abstract, and intentionally so, but on a philosophical level, it's closely related to Shannon's signal-theory definition of "information". - Datastore: Amount of information that is accessible through one or more access methods. - User: Entity that may (try to) access information in a datastore. Note that no assumption is made that the user is animal, vegetable or mineral. - Access method: Well-defined series of operations that will cause information known to a datastore to also be known to the user. - Site: Entity that hosts all or part of a datastore, and makes it available through one or more access methods. A site may in various contexts be a machine, a datacenter, a network of datacenters, or a single device. 2. Dimensions of classification 2.1 Uniqueness and scope Some information systems are global, in the sense that only one can sensibly exist in the world. Others are inherently local, in that each locality, site or even box will run its own information store, independent of all others. The following terms are suggested: - Global datastore: A datastore that there can be only one of in the world. The world itself is a prime example; the public telephone system's number assignments is another. - Local datastore: A class of datastore of which multiple instances can exist, each with information relevant to that particular datastore, with no need for coordination between them. ((( better term needed ))) - Centralized datastore: A datastore where all access to data has to pass through some single point of control (site). - Distributed datastore: A datastore that is not centralized. - Replicated datastore: A distributed datastore where all sites have the same information - Cooperative datastore: A distributed datastore where not all sites have all the information, but where mechanisms exist to get the info to the requester, even when it is not available to the site originally asked draft-alvestrand-lang-tags-v2-01.txt [Page 2] Definitions for talking about directories Harald Alvestrand draft-alvestrand-directory-defs-00.txt Expires April 2000 2.2 Search, Lookup, Query and Notify A central consideration when describing datastores is the types of method they offer to find information. The chief classifications are: - Lookup datastores require the user to know or guess some exact value before asking for information, sometimes called a "lookup key" and sometimes called a "name". They usually return a single piece of information as a response. - Search datastores require the user to know some approximate value of some information. They usually return zero, one or more responses that match the information supplied according to some algorithm. An orthogonal dimension has to do with time: - Query datastores will answer a request with a response, and once that is over with, will do nothing more. - Notify datastores will get a request from an user to have information returned at some later time when it becomes available, current or whatever, and will respond at that time with a notification that information is available. - Subscription datastores are like notify datastores, but will transfer the actual information when available. 2.3 Consistency models Consistency (or the lack thereof) is a property of distributed datastores; for this particular discussion, we ignore the subject of semantically inconsistent data (such as an assertion that a man is blind and has a valid driver's license), and focus on the problem of consistency where inconsistency is defined as having the same request, using the same credentials, be answered with different data at different sites. Distributed datastores may have: - Strict consistency, where the problem above never arises - Strict internal consistency, where the replies always reflect a consistent picture of the total datastore, but some sites may reflect an earlier version of the datastore than others - Loose, converging consistency, where different parts of the datastore may be updated at different times as seen from a single site, but the process is designed in such a way that if one stops making changes to the datastore, all sites will sooner or later present the same information - Inconsistency, where no guarantee can be made whatsoever One interesting variant is subset consistency, where the system is consistent (according to one of the definitions above), but not all draft-alvestrand-lang-tags-v2-01.txt [Page 3] Definitions for talking about directories Harald Alvestrand draft-alvestrand-directory-defs-00.txt Expires April 2000 questions will be answered at all sites; possibly because different sites have different policies on what they make available (NetNews), or because different sites only need different subsets of the "whole picture" (BGP). 2.4 Security models It's harder to describe security models in a few sentences than other properties of information systems. Some thoughts, though: On trust in information: Why do we trust a piece of information to be correct? - Because it's in the datastore (and therefore must have been authorized). This is perimeter (or Eggshell) integrity. - Because it contains internal integrity checks, usually involving digital signatures by verifiable identities This is item integrity; the granularity of the integrity and the ability to do integrity checks on the relationships between objects is extremely important and extremely hard to get right, as is establishing the root of the trust chain. - Because it fits other available information, and causes the right things to happen when I use it. This is hopeful integrity. Which integrity model to choose is a matter of evaluating the cost of implementing the integrity, the cost of having the integrity break on you, and the impact of cost of doing business. On access to information, the usual categories apply: - Open access: Anyone can get the information. - Access because of what you are: Limited to "same network", "physically present" or "resolvable DNS name" - Access because of who you are (in theory): username/password, certificatesą.. These are then backed up by a layer specifying what the identity you have proven yourself to be has access to - Access because of what you have: hardware tokens, smartcards, certificates, capability keysą. In this case, access is given to all who can present that credential, without caring about their identity. The most common approaches are identity-based and open access. 2.5 Update models Two words about update models: draft-alvestrand-lang-tags-v2-01.txt [Page 4] Definitions for talking about directories Harald Alvestrand draft-alvestrand-directory-defs-00.txt Expires April 2000 - Read-only datastores have no standard means of changing the information in them. This is usually accomplished through some other interface than the standard interface. - Read-mostly datastores are designed based on a theory that reads will greatly outnumber updates; this may, for instance, be reflected in relatively slow consistency-updating protocols. - Read-write datastores assume that the updates and the read operations are of the same order of magnitude. 3. Classification of some real systems 3.1 The Domain Name System The DNS is a global lookup datastore with loose, converging consistency and query capability only. It is either strictly read-only or read-mostly (with Dynamic DNS), has an open access model, and mainy perimeter integrity (some would say hopeful integrity). DNSSEC aims to give it item integrity. If one opens up the box and looks at the relationship between primary and secondary nameservers, that can be seen as a limited form of notify capability, but this is not available to end-users of the total system. 3.2 The (imagined) X.500 Global Directory X.500 was intended to be a global search datastore with loose, converging consistency. It was intended to be read-mostly, perimeter secure and query-capable. 3.3 The Global BGP Routing Information Database The Global or top-level BGP routing information database is a global read-write datastore with loose, converging subset consistency (not all routes are carried everywhere) and very limited integrity control, mostly intended to be perimeter integrity based on "access control based on what you are". 3.4 The NetNews system NetNews is a global read-write datastore with loose (non-converging) subset consistency (not all sites carry all articles, and article retention times differ). Between sites it offers subscription capability; to users it offers both search and lookup functionality. 3.5 SNMP MIBs An SNMP agent can be thought of as a local, centralized datastore offering lookup functionalty. With SNMPv3, it offers all kinds of access models, but mostly "access because of what you have" seems popular. draft-alvestrand-lang-tags-v2-01.txt [Page 5] Definitions for talking about directories Harald Alvestrand draft-alvestrand-directory-defs-00.txt Expires April 2000 3.6 The MBONE MBONE can be thought of as a highly transient, read-write datastore with subscription capability. 4. Security Considerations Security is a very relevant question when considering information access systems. Some issues to consider are: - Controlled access to information - Controlled rights to update information - Protection of the information path from provider to consumer - With personal information, privacy issues - Interactions between multiple ways to access the same information 5. Character set considerations @ 6. Acknowledgements 7. Author's Address Harald Tveit Alvestrand EDB Maxware Pirsenteret N-7462 TRONDHEIM NORWAY EMail: Harald.Alvestrand@maxware.no Phone: +47 73 54 57 97 @ References Appendix A: List of language tags draft-alvestrand-lang-tags-v2-01.txt [Page 6]