Quiz questions

1
Lecture 9

• Quiz questions
• Epidemic communication protocols
• Naming services

2
Quiz sample question

• Discuss the advantages and drawbacks of a server
  design using non-blocking I/O, a state machine
  and a single thread with a design using blocking
  I/O multi-threading.

3
Quiz sample question

• The C language has a construct called union in
  which a field of a record can hold one of several
  alternatives.
• For example the following piece of code declares
  a new union type union_def. Variables of this
  type can hold either one of an integer, a float
  or a character.
• union union_def int a float b char c //
  define the type
• union union_def union_var // define the
  variable
• union_var.b99.99 or // initialize the
  variable
• union_var.a34
• At runtime there is no clear cut way for the
  runtime system to tell what datatype is stored in
  a variable declared as an union.
• Does this language feature have any implication
  for remote procedure calls? Explain your answer.

4
Q10

• Protocol clients sends 16 bytes code. Server
  answers 0/1 encoded in one byte.
• Question Is the following code correct?
• public class ccServer
• private static final int BUFSIZE 16
  // Size of receive buffer
• public static void main(String args) throws
  IOException
• int servPort Integer.parseInt(args0)
• // Create a server socket and buffer to
  accept client connections
• ServerSocket servSock new
  ServerSocket(servPort)
• byte byteBuffer new byteBUFSIZE
  // Receive buffer
• for () // Run forever, accepting and
  servicing connections
• Socket clntSock servSock.accept()
  // Get client connection
• InputStream in clntSock.getInputStr
  eam()
• OutputStream out clntSock.getOutputS
  tream()
• // Receive until client closes connection,
  indicated by -1 return
• in.read(byteBuffer))

5

• What are the pros and cons of using epidemic
  protocols for disseminating information?

6
Last time Anti-Entropy Protocols

• A node P selects another node Q from the system
  at random.
• Push P only sends its updates to Q
• Pull P only retrieves updates from Q
• Push-Pull P and Q exchange mutual updates (after
  which they hold the same information).
• Observation for push-pull it takes O(log(N))
  rounds
• to disseminate updates to all N nodes
• one round every node as taken the initiative to
  start one exchange.
• Main properties
• Reliability a node failures do not impact the
  protocol
• Dissemination time effort, scales well with the
  number of nodes

7
Last time Gossiping

• Basic model A node S having an update to report,
  contacts other randomly chosen servers.
• Termination decision
• If the contacted node already has the update S
  stops contacting other nodes with probability
  1/k.
• P the share of nodes that have not been reached
• P e -(k1)(1-p)

ln(P)
8
Last time Example applications

• Data dissemination in p2p, wireless sensor
  networks, clusters
• Spreading updates
• E.g., disconnected replicated list maintenance
  Demers et al., Epidemic algorithms for
  replicated database maintenance. SOSP87
• Membership protocols
• e.g., Amazon Dynamo service DeCandia et. al,
  Dynamo Amazons Highly Available Key-value
  Store, SOSP07
• Various p2p networks (e.g., Tribler)
• Data aggregation
• The problem compute the average value for a
  large set of sensors
• Let every node i maintain a variable xi. When two
  nodes gossip, they each reset their variable to
• xi, xk ?(xi xk)/2
• Result in the end each node will have computed
  the average
• avg sum(xi))/N.

9
Advantages of epidemic techniques

• Probabilistic model. Rigorous mathematical
  underpinnings. Good framework for reasoning about
  the spread of information through a system over
  time.
• A-synchronous communication pattern. Operate in a
  'fire-and -forget' mode, where, even if the
  initial sender fails, surviving nodes will
  receive the update.
• Autonomous actions. Enable you to take actions
  based on the data you received without the need
  for additional communication to reach agreement
  with your partners you can take decisions
  autonomously.
• Robust with respect to message loss node
  failures. Once a message has been received by at
  least one of your peers it is almost impossible
  to prevent the spread of the information through
  the system.

10
Naming Systems Distributed Hash Tables
11
Naming systems

• Functionality
• Map names ? access points (addresses)
• Names are used to denote entities in a
  distributed system.
• To operate on an entity, we need to access it at
  an access point (address).
• Note A location-independent name for an entity
  E, is independent from the addresses of the
  access points offered by E.

12
Names are valuable!

• NYT, August00

13
Naming systems

• Functionality
• Map names ? access points (addresses)
• Main challenge scaling
• of names, clients
• geographical distribution,
• management

14
Terminology

• Names
• Human friendly vs. arbitrary (random strings)
• Names vs. identifiers
• Identifiers have three properties
• refer to at most one entry
• each entity is referred by at most one identifier
• always refers to the same entity
• Name space
• Flat vs. hierarchical

15
Next

• Naming system implementations
• Flat namespace
• Distributed Hash Tables
• Hierarchical Namespace
• DNS

16
Strawman (1)

• Why not centralize?

17
Strawman (1)

• Why not centralize?
• Single point of failure
• High latency
• Distant centralized database
• Scalability bottleneck
• Traffic volume
• Management Single point of update

18
Strawman (2)

• Why not use /etc/hosts?

19
Strawman (2)

• Why not use /etc/hosts?
• Original Name to Address Mapping
• Flat namespace
• /etc/hosts
• SRI kept main copy
• Downloaded regularly
• Count of hosts was increasing machine per domain
  ? machine per user
• Many more downloads
• Many more updates
• Still a scalability bottleneck

20
Implementation options Flat namespace

• Problem Given an essentially unstructured name
  how can we locate its associated address?
• Possible designs
• Simple solutions (broadcasting, forwarding
  pointers)
• Home-based approaches
• Distributed Hash Tables (structured P2P)

21
Flat namespaces simple solutions

• Broadcasting Simply broadcast the ID, requesting
  the entity to return its current address.
• Can never scale beyond local-area networks (think
  of ARP/RARP)
• Requires all processes to listen to incoming
  location requests
• Forwarding pointers Each time an entity moves,
  it leaves behind a pointer telling where it has
  gone to.
• Update a clients reference as soon as present
  location has been found
• Geographical scalability problems
• Long chains are not fault tolerant
• Increased network latency at dereferencing

22

• Functionality to implement
• Map names ? access points (addresses)
• Similar to a hash-table
• Manage (huge) list of (name, access point) pairs
• Put (key, value)
• Lookup (key) ? value
• Key idea partitioning.
• Allocate parts of the list to different nodes

23
Partition Solution Consistent hashing

• Consistent hashing
• the output range of a hash function is treated as
  a fixed circular space or ring.

0
128
Circular ID Space
24
Partition Solution Consistent hashing

• Mapping keys to nodes
• Advantages incremental scalability, load
  balancing

N10
Key ID Node ID
K5, K10
N100
K99
Circular ID Space
N32
K11, K30
N80
K65, K70
N60
K33, K40, K52
25
Consistent hashing

• How do store lookup work?

N10
Key ID Node ID
K5, K10
N100
K99
N32
K11, K30
N80
K65, K70
N60
K33, K40, K52
26
Consistent Hashing -- Summary

• Mechanism
• Nodes get an identity by hashing their IP
  address, keys are also hashed into same space
• A key with id (hashed into) k, is assigned to
  first node whose hashed id is equal or follows k,
  in circular space successor(k)
• Advantage
• Incremental scalability, Balanced Distribution,
• Theoretical results
• N number of nodes, K number of keys in the
  system
• With high probability Each node is responsible
  for at most (1?)K/N keys
• With high probability Joining or leaving of a
  node relocates O(K/N) keys (and only to or from
  the responsible node)

27
Extra consistent hashing tricks

• Virtual Nodes Each physical node can be
  responsible for multiple virtual nodes.
• Advantage load balancing
• Dealing with heterogeneity The number of virtual
  nodes that a node is responsible for can decided
  based on its capacity, accounting for
  heterogeneity in the physical infrastructure.
• If a node becomes unavailable the load handled by
  this node is evenly dispersed across the
  remaining available nodes.
• When a node becomes available again, the newly
  available node accepts a roughly equivalent
  amount of load from each of the other available
  nodes.