Chapter 5- Naming

Mulugeta A.

1
Naming
 Names play a critical role in all computer systems
 Used to share resources
 Used to refer to locations
 To uniquely identify entities in a distributed system, and more
 A name in a distributed system is a string of bits or characters that
is used to refer to an entity.
 hosts, printers, disks, files, processes, web-pages
 An important issue in naming is that a name can be resolved to the
entity it refers to.
 Name resolution system thus allows a process to access the named entity.
 In distributed system, the implementation of a naming system is often
distributed across multiple machines.
 How distribution is done has impact on the performance

2
Cont...
 Entities can be operated on.
 E.g. Printer provides operations for printing, pausing, cancelling etc.
 An entity such as a network connection may provide operations for sending and receiving
data, setting quality-of-service parameters, requesting the status, and so forth.
 To operate on an entity, it is necessary to access it, for which we need an access point.
 An access point is yet another, but special, kind of entity in a distributed system.
 The name of an access point is called address
 E.g. A host running a specific server, its address formed by the combination of IP address and port
number
 An entity may have multiple access points, and its access point may change
 Thus, the address of an access point should not be used to name the entity
 For example, when a mobile computer moves to another location, it is often assigned a
different IP address than the one it had before.
 Likewise, when a person moves to another city or country, it is often necessary to change
telephone numbers as well.
 In a similar fashion, changing jobs or Internet Service Providers, means changing your e-
mail address.

3
Cont...
 Therefore, what we need is a name for an entity that is independent
from its addresses
 i.e., a location-independent name
 A location-independent name for an entity E , is independent from the
addresses of the access points offered by E .
 Identifier
 Identifier is a name having the following properties:
 P1: Each identifier refers to at most one entity
 P2: Each entity is referred to by at most one identifier
 P3: An identifier always refers to the same entity (prohibits
reusing an identifier)

4
The Central Theme of Naming
 How to resolve names and identifiers to addresses
 In principle, a naming system maintains a name-to-
address binding in the form of mapping table
 However, a centralized table in a large network is not going to
work
 The name resolution as well as the table is often
distributed across multiple machines
 E.g., www.xx.edu.et is divided into multiple parts, and several
DNS servers are used during the resolution
 Will discuss the resolution of the following names
 Flat names, structured names, attribute-based names

5
Flat Names (Identifiers)
 An identifier is often a string of random bits
 Does not contain any information on how to locate the access
point of its associated entity
 Problem:- given an essentially unstructured name (e.g., an
identifier), how can we locate its associated access point?
 Simple solutions (broadcasting)
 Home-based approaches
 Distributed Hash Tables (structured P2P)
 Hierarchical location service

6
Simple solutions
 Two simple solutions to locate the entity given an
identifier
 Broadcasting and multicasting (e.g., ARP)
 Broadcast the ID, requesting the entity to return its current address.
 Can never scale beyond local-area networks
 Requires all processes to listen to incoming location requests
 Forwarding pointers
 A popular approach to locate mobile entities
 When an entity moves, it leaves a pointer to where it went
 Update a client’s reference when present location is found
 Advantage:
 Dereferencing can be made transparent to client by following the
pointer chain

7
Forwarding Pointers
 Disadvantages
 Geographical scalability problems:
 Chain can be very long for highly mobile entities
 Long chains are not fault tolerant
 High latency when dereferencing
 Need chain reduction mechanisms
 Update client's reference when the most recent location is found

8
Home-based solution
 Let the home keep track where the entity is.
 Entity’s home address is registered at a naming service.
 In practice, the home location is often chosen to be the place where an
entity was created.
 The home registers the foreign address of the entity.
 Clients always contact the home first, and then continue with
the foreign location.

9
 Home-Based Approaches

Figure 5-3. The principle of Mobile IP.

10
Problems With Home-Based Approaches
 Home address has to be supported for entity’s lifetime.
 Home address is fixed.
 Unnecessary burden when the entity permanently moves.
 Poor geographical scalability.
 Entity may be next to client.

11
Distributed Hash Table (DHT)- Chord
 In DHT-based system,
 Each node has an m-bit random identifier
 Each entity has an m-bit random key
 An entity with key k is located on a node with the smallest identifier
that satisfies id >= k, denoted as succ(k)
 The main issue in DHT-based systems is to efficiently resolve
a key k to the address of succ(k).
 Two approaches:
 linear approach and finger table
 In linear approach, let each node p keep track of the successor
succ(p+ 1) as well as its predecessor pred(p)
 whenever a node p receives a request to resolve key k, it will simply
forward the request to one of its two neighbours
12
Finger table
 Each node p maintains a finger table FTp[] with at most m
entries:
 FTp[i] = succ (p+ 2i−1)
 Note: FTp[i] points to the first node succeeding p by at least 2 i−1
 To look up a key k, node p forwards the request to node q
with index j in the figure table satisfying
 q= FTp[ j] ≤ k< FTp[ j+1]
 If p < k < FTp[1], the request is also forwarded to FTp[1].

13
Distributed Hash Tables

Figure 5-4. Resolving key 26 from node 1 and key 12 from node 28 in a Chord system.
14
Exploiting Network Proximity
 Problem:
 The logical organization of nodes in the overlay may lead to
erratic message transfers in the underlying Internet
 Node k and node succ(k +1) maybe very far apart.
 Solutions:
 Topology-aware node assignment: When assigning an ID to a
node, make sure that nodes close in the ID space are also close
in the network.
 Proximity routing: Maintain more than one possible successor,
and forward to the closest.
 Proximity neighbour selection: When there is a choice of
selecting who your neighbour will be, pick the closest one.

15
Hierarchical Location Services
 The basic idea is to build a large-scale search tree for which the underlying
network is divided into hierarchical domains.
 Each domain is represented by a separate directory node.
 Leaf domains typically correspond to a local-area network or a cell.
 The root (directory) node knows all the entities.
 Each entity currently in a domain D is represented by a location record in the
directory node dir(D) which is the entity’s current address or a pointer.

16
Hierarchical Approaches- organization
 The address of an entity E is stored in a leaf or intermediate node.
 Intermediate nodes contain a pointer to a child if the sub-tree rooted at the child
stores an address of the entity.
 An entity may have multiple addresses (e.g., if it is replicated).

An example of storing information of an entity having two addresses in different leaf domains.
17
Hierarchical Approaches - Lookup Operation
 Start lookup at local leaf node.
 If node knows about entity E, follow downward pointer, otherwise go up.
 Upward lookup always stops at root.

Looking up a location in a hierarchically organized location service.

18
Hierarchical Approaches – Insert
operation
 Consider an entity E that has created a replica in leaf domain D for which it
needs to insert its address.
 [A]. An insert request is forwarded to the first node that knows about entity E.
 [B]. A chain of forwarding pointers to the leaf node is created.

19
Structured Naming
 Flat names are not convenient for humans to use
 As a result, naming systems often support structured
names that
 Are composed from simple, human-readable names
 E.g., file names, Internet domain names
 Structured names are often organized into what is called a
name space
 Holds a collection of valid names recognized by a particular
service.
 Examples: Phone numbers, DNS, URL etc
 It is represented by a labelled, directed graph with two types of
nodes, leaf node and directory node
20
Con’t
 A leaf node represents a named entity and has the property
that it has no outgoing edges.
 A leaf node generally stores information on the entity it is
representing–for example, its address–so that a client can
access it.
 A directory node has a number of outgoing edges, each
labeled with.
 Each node in a naming graph is considered as yet another entity in
a distributed system, and, in particular, has an associated identifier.
 A directory node stores a table in which an outgoing edge is
represented as a pair (node identifier, edge label). Such a
table is called a directory table.
21
Con’t
 A general naming graph with a single root node.

Note:- A directory node contains a (directory) table of (node identifier, edge label) pairs.
22
Name Resolution
 Name spaces offer a convenient mechanism for storing and retrieving information
about entities by means of names.
 More generally, given a path name, it should be possible to look up any information
stored in the node referred to by that name.
 The process of looking up a name is called name resolution.
 Name resolution refers to the process of looking up a name.
 To resolve a name we need a directory node.
 Problem:
 How do we actually find that (initial) node?
 Solution:
 Closure Mechanism: the mechanism to select the implicit context from which to start name
resolution.
 Knowing how and where to start name resolution is generally referred to as a closure
mechanism.
 Examples:
- www.xx.edu.et: start at a DNS name server
- /home/steen/mbox: start at the local file server
 Start from well know root directory, or start from home directory
23
Linking and Mounting
 Aliases are commonly used in a name space
 An alias is another name for the same entity.
 In naming graphs, there are basically two different ways to implement an alias :
hard link and a symbolic link
 Hard Link
 Multiple absolute path names refer to the same node in a naming graph.
 names resolved by following a specific path in a naming graph from one node to
another.
 For example in the above diagram, in which node n5 can be referred to by two
different path names. Both path names /keys and /home/steen/keys
 Symbolic Link
 It represent an entity by a leaf node, say N, but instead of storing the address or
state of that entity, the node stores an absolute path name.
 For example, as it can be shown in the following figure the path name
/home/steen/keys, which refers to a node containing the absolute path name
/keys, is a symbolic link to node n5.
24
Linking and Mounting

The concept of a symbolic link explained in a naming graph.

25
Linking and Mounting
 A collection of name spaces can be distributed across
different machines.
 Each name space implemented by separate servers
 To mount a foreign name space in a distributed system
requires at least the following information:
 The name of an access protocol.
 The name of the server.
 The name of the mounting point in the foreign name space.

26
Linking and Mounting
 The organization of a file system on the client machine is partly shown in the
following figure.
 The root directory has a number of user-defined entries, including a subdirectory
called /remote.
 This subdirectory is intended to include mount points for foreign name spaces
such as the user’s home directory at VU University.
 To this end, a directory node named /remote/vu is used to store the URL
nfs ://its.cs.vu.nl/home/steen.
 Now consider the name /remote/vu/mbox.
 This name is resolved by starting in the root directory on the client’s machine and
continues until the node /remote/vu is reached.
 The process of name resolution then continues by returning the URL nfs
://its.cs.vu.nl/home/steen, in turn leading the client machine to contact the file server
its.cs.vu.nl by means of the NFS protocol, and to subsequently access directory
/home/steen.
 Name resolution can then be continued by reading the file named mbox in that directory,
after which the resolution process stops.
27
Linking and Mounting
 Mounting remote name spaces through a specific access protocol.

28
Implementation of a Name Space
 A name space is often implemented by name servers
 In LAN, a single name server is often good enough
 In large-scale distributed system, the implementation of a name space is
often distributed over multiple name servers
 A name space for large-scale distributed systems is often
organized hierarchically
 Global layer
 often stable, represents organizations or groups of organizations
 Administrational layer
 Represents groups of entities in a single organization or admin. unit
 Managerial layer
 Nodes often change frequently, e.g., hosts in a local network
 May be managed by system administrators or end users
29
 Name Space Distribution

An example partitioning of the DNS name space, including Internet-accessible files, into three layers.

30
Name Space Distribution

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.

31
Implementation of Name Resolution
 Assume that the (absolute) path name
 root: <nl, vu, cs, ftp> is to be resolved.
 Using URL notation, this path would correspond to ftp://cs.vu.nl
 Two ways to implement name resolution
 Iterative name resolution
 Hands over the complete name to the root name server.
 Intermediate result(i.e., the address of the next name server in the
hierarchy) is returned back to the client’s name resolver.
 The resolution process continues iteratively until the complete path is
resolved.
 Recursive name resolution
 Intermediate results are passed to next name server in the hierarchy.
 This process continues recursively until the complete path is resolved.
32
Implementation of Name Resolution
 The principle of iterative name resolution.

33
Implementation of Name Resolution
The principle of recursive name resolution.

34
Iterative Vs Recursive Resolution

 Recursive name resolution of <nl, vu, cs, ftp>.

 Name servers cache intermediate results for subsequent lookups.
35
Iterative Vs Recursive Resolution

The comparison between recursive and iterative name resolution with respect to
communication costs.

36
Domain name system
 DNS is one of the largest distributed naming services in use today.
 Used for looking up host addresses and mail servers.
 The DNS name space is hierarchically organized as a rooted tree.
 A sub-tree is called a domain;
 a path name to its root node is called domain name.
 In practice, the DNS namespace can be divided into a global layer and
administrational layer.
 The managerial layer, which is generally formed by local file systems, is
formally not part of DNS.
 A node is the DNS name space often represents several entities at the same
time, e.g., a domain and a zone.

 The content of a node is formed by a collection of resource records.

37
The DNS Name Space
 The most important types of resource records forming the contents of nodes
in the DNS name space.

38
Attribute Based Naming
 Attribute-based naming name and lookup entities by
means of their attributes
 Also known directory service (yellow page)
 Attribute-based naming
 Each entity is associated with a collection of attributes
 The naming system provides one or multiple entities that match
a user’s description
 Problem
 Lookup operations can be extremely expensive,
 As it demands to match requested attribute values, against actual attribute
values
 Done by inspecting all entities (in principle)

39
 End of Lecture 5

40