Chapter 5 Naming (2)

Transcription

1 DISTRIBUTED SYSTEMS Principles and Paradigms Second Edition ANDREW S. TANENBAUM MAARTEN VAN STEEN Chapter 5 Naming (2)

2 Plan Definitions and uses Types of naming Flat naming Structured naming Attribute-based naming

3 Structured Naming Flat naming convenient for machines, not so for humans Support names composed from simple, human-readable names /users/marius/public_html/

4 Name Spaces Can be represented as a directed graph Leaf node represents named entity No outgoing edges Directory nodes store information on other nodes Outgoing edge per referenced identifier

5 Example: Linux proc file system Pseudo filesystem Used as an interface to kernel datastructures Rooted at /proc E.g., /proc/[number]/ Numerical subdirectory for each running process /proc/filesystems

6 Name Resolution Given a path name, retrieve information of entity referenced by that name N:<label1,, labeln> E.g., root:<dk, au, daimi, www> -> Start at node N, follow label1, labeln Need closure mechanism, i.e., knowledge of where to start lookup

7 Name Space Implementation Need distributed solution for large-scale networks Logical layers Global layer Root node and nodes logically close to Stable assignment Administrational layer Managed within single organization Managerial layer Maintained also by users Unstable assignment

8 Name Space Distribution Figure An example partitioning of the DNS name space, including Internet-accessible files, into three layers.

9 Domain Name System (DNS) Distributed directory service Hierarchical name space Each level separated by. Analogous to / separator in file systems One global root Replicated across 13 root servers There have been Denial of Service (DoS) attacks on these root servers, none real successful Because of caching, queries to root servers relatively rare DNS maybe only global directory/name service???

10 Name Space Distribution Figure A comparison between name servers for implementing nodes from a large-scale name space partitioned into a global layer, an administrational layer, and a managerial layer.

11 DNS is simple but powerful Only one type of query Query(domain name, RR type) Resource Record (RR) type is like an attribute type Answer(values, additional RRs) Limited number of RR types Hard to make new RR types Not for technical reasons Rather because each requires global agreement

12

13 Important DNS RR types NS: Points to next Name Server down the tree A: Contains the IP address AAAA for IPv6 MX: Contains the name of the mail server Service-oriented RR types SRV: Contains addresses and ports of services on servers One way to learn what port number to use NAPTR: Essentially a generalized mapping from one name space (i.e. phone numbers) to another (i.e. SIP URL)

14 The DNS Name Space Figure The most important types of resource records forming the contents of nodes in the DNS name space.

15 DNS Tree Structure. NS RR pointers dk. com. edu. us. au.dk. iha.dk. itu.dk daimi.au.dk. nat.au.dk. A svn.daimi.au.dk A

16 Primary and secondary servers au.dk. NS RRs point to both primary and secondary servers daimi.au.dk. RRs are initially configured into primary server Primary server replicates RRs onto secondary servers periodically (updates are incremental)

17 Implementation of Name Resolution Figure The principle of iterative name resolution.

18 Implementation of Name Resolution Figure The principle of recursive name resolution.

19 Implementation of Name Resolution Figure Recursive name resolution of <nl, vu, cs, ftp>. Name servers cache intermediate results for subsequent lookups.

20 Example: The Domain Name System Figure The comparison between recursive and iterative name resolution with respect to communication costs.

21 DNS Implementation Figure An excerpt from the DNS database for the zone cs.vu.nl.

22 DNS Implementation Figure An excerpt from the DNS database for the zone cs.vu.nl.

23 Decentralized DNS Implementation Can we use DHTs to implement DNS? Hash each DNS name to k Make node with id lookup(k) responsible for maintaining records for the name Good Scalability via lookup performance Semantic-free mapping Bad Looses structure of DNS names Scalability via caching?

24 Decentralized DNS Implementation Prefix of length 0 shared Pastry DHT Routing table Leaf set Nearest neighbours in id space CoDoNS Based on Pastry Replicate DNS entry with id k to nodes that share a prefix with k Length of shared prefix parameterizes algorithm Prefix of length 4 shared

25 Attribute-Based Naming LDAP is another popular distributed directory service Richer and more general than DNS Has generalized attribute/value scheme Can search on attribute, not just name Simpler and more efficient than a full relational database Not a global directory service, though namespace is global Its predecessor, X.500, was meant to be But local LDAP services can point to each other Commonly used for personnel RR databases, subscriber databases

26 Hierarchical Implementations: LDAP (1) Figure A simple example of an LDAP directory entry using LDAP naming conventions.

27 Hierarchical Implementations: LDAP (2) Figure (a) Part of a directory information tree.

28 Hierarchical Implementations: LDAP (3) Figure (b) Two directory entries having Host_Name as RDN.

29 Searching in LDAP May specify criterias over attributes E.g., search( &(C=NL)(O=Vrije Universiteit)(OU=*)(CN=Main server) May be expensive Need to iterative over all OU at O in this case

30 Decentralized Implementations Use DHTs for attribute-based naming and search

31 Mapping to Distributed Hash Tables Figure (a) A general description of a resource. (b) Its representation as an AVTree.

32 Mapping to Distributed Hash Tables Need to transform tree into keys Hash every (sub)path in tree to a key hash(type-book) hash(type-book-author) hash(genre-fantasy) Nodes responsible for a key will store pointers to actual ressource I.e., LoTR May support range queries by having multiple servers for each node

33 Mapping to Distributed Hash Tables Figure (a) The resource description of a query. (b) Its representation as an AVTree.

34 Summary Naming is fundamental to distributed systems Different types of names may be used Flat naming E.g., DHT Structured naming E.g, DNS Attribute-based naming E.g., LDAP