CTIS 490 DISTRIBUTED SYSTEMS

1
CTIS 490DISTRIBUTED SYSTEMS

• WEEK 3
• DISTRIBUTED SYSTEMS
• PROCESSES

2
INTRODUCTION

• Processes play a fundamental role in distributed
  systems since they form a basis for communication
  between different machines.
• An important issue is how processes are
  internally organized and whether they support
  multiple threads of control.
• Threads in distributed systems are particularly
  useful to continue using the CPU when a blocking
  I/O operation is performed.

3
MULTITHREADING

• The ability of an operating system to execute
  different parts of a program, i.e. threads, on a
  single processor, is called multithreading.
• Multithreading also exploits parallelism when
  executing the program on a multiprocessor system.
• In that case, each thread is assigned to
  different CPU while shared data are stored in
  main memory.
• In distributed systems, multithreading is used in
  both client and server process design.

4
MULTITHREADED CLIENTS

• To establish a high degree of distribution
  transparency, distributed systems that operate in
  wide-area networks may need to conceal long
  inter-process message propagation times.
• The usual way to hide communication latencies is
  to initiate communication and immediately proceed
  with something else.
• A typical usage of multithreading is in Web
  browsers.

5
WEB BROWSER

• The Web browser has to set up a TCP/IP
  connection, read incoming data, and pass it to a
  display component.
• Setting up a connection as well as reading
  incoming data are blocking operations.
• To hide communication latencies, browser starts
  displaying data while it is still coming in.
• While the text is made available to the user,
  including the facilities for scrolling and such,
  the browser continues with fetching other files
  that make up the page such as the images.
• Developing the browser as a multithreaded client
  simplifies the design.

6
MULTITHREADED SERVERS

• Multithreading in servers simplifies code design
  and makes it much easier to develop servers that
  exploits parallelism to attain high performance.
• To understand the benefits of threads for writing
  server code, consider the organization of a file
  server that occasionally has to block waiting for
  the disk.
• The file server normally waits for an incoming
  request for a file operation, subsequently
  carries out the request, then sends back the
  reply.

7
MULTITHREADED SERVERS

• One thread, the dispatcher, reads incoming
  requests for a file operation.
• The server chooses an idle worker thread and
  hands it the request.

8
CLIENT DESIGN

• A major task of client machines is to provide the
  means for users to interact with remote servers.
• There are basically two ways in which this
  interaction can be supported.

9
CLIENT DESIGN

• For each remote service the client machine will
  have a separate counterpart that can contact the
  service over the network.
• For example, an agenda running on a users PDA
  that needs to synchronize with remote shared
  agenda.
• In this case, an application-level protocol will
  handle the synchronization.

10
CLIENT DESIGN

• Provide access to remote services by only
  offering a convenient user interface.
• This means tha client machine is used only as a
  terminal with no need for local storage, leading
  to an application neutral solution.

11
CLIENT DESIGN

• Client software comprises more than just user
  interfaces. In many cases, parts of the
  processing and data level in a client-server
  application are executed on a client side.
• Client software also comprises components for
  achieving distribution transparency i.e. The user
  should not be aware that it is communicating with
  remote processes.
• For example, client software can be configured to
  handle failure transparency. It can attempt to
  connect to a server repeatedly or try another
  server after several attempts.
• Client software returns data it had cashed during
  a previous session, as is sometimes done by Web
  browsers that fail connect to a server.

12
SERVER DESIGN

• A server is a process which implements a specific
  service on behalf of a collection of clients.
• There are two ways to organize servers
• Iterative server the server itself handles the
  request and returns a response to the requesting
  client.
• Concurrent server the server does not handle
  the request itself, but passes it to a separate
  thread or process, after which it immediately
  waits for the next incoming request.

13
SERVER DESIGN

• How do clients know where to contact for service?
• Clients send requests to an end point, also
  called a port, at the machine where server is
  running.
• Each server listens to a specific end point.
• One approach is to globally assign end points for
  well-known services.

14
SERVER DESIGN

• For example, servers that handle Internet FTP
  requests always listen to TCP port 21.
• Likewise, an HTTP server always listens to TCP
  port 80.
• These end points have been assigned by the
  Internet Assigned Numbers Authority (IANA).
• With assigned end points, the client only needs
  to find the network address of the machine where
  the server is running.

15
SERVER DESIGN

• There are many services that do not require a
  pre-assigned end point.
• For example, a time-of-the day server may use an
  end point that is dynamically assigned to it by
  its local operating system.
• In that case, the client will first have to look
  up the end point.
• One solution is to have a special daemon running
  on each machine that runs servers which listens
  to a well-known point.

16
SERVER DESIGN
17
SERVER DESIGN

• One final design issue is whether or not the
  server is stateless or stateful.
• Stateless server does not keep information on the
  state of its clients.
• For example, a Web server is stateless server.
• Stateful server maintains persistent information
  on its clients.
• For example, a file server maintains a table
  containing (client,file) entries.
• Such a table allows the server to keep track of
  which client has the update permissions on which
  file.

18
SERVER DESIGN

• When the server cannot maintain state, a common
  solution is to let the client send along
  additional information on its previous accesses.
• In the case of Web, this information is
  transparently stored by the clients browser in
  what is called a cookie.
• A cookie is small piece of data containing
  client-specific information that is of interest
  to the Web server.

19
SERVER CLUSTERS

• A server cluster is a collection of machines
  connected through a network where each machine
  runs one or more servers.
• The server clusters that we consider here are the
  ones in which machines are connected through a
  LAN offering high bandwidth.
• In most cases, a server cluster is organized into
  three tiers.

20
SERVER CLUSTERS

• First tier consists of a switch through which the
  client request are routed.
• Second tier, contains servers dedicated to
  application processing. In cluster computing,
  these are typically servers running on
  high-performance hardware dedicated delivering
  compute power.
• Third tier contains data-processing servers,
  running file and database servers. Depending on
  the usage of the server cluster, these servers
  may be running specialized machines, configured
  for high-speed disk access and having large data
  cashes.