Olivier H. Martin (1)

1

Multicast, IPv6, CDN, Web Caches HTTP

• with special thanks to Cisco (Pei Cao), SUN
  (Alain Durand)
• Lecture 4 presented at the 26th International
  Nathiagali Summer College on Physics and
  Contemporary Needs, 25th June 14th July,
  Nathiagali, Pakistan
• Olivier H. Martin
• CERN - IT Division
• June 2001
• Olivier.Martin_at_cern.ch

2
Native Multicast
3
Native Multicast
4
(No Transcript)
5
Multicast

• MBONE
• DVMRP tunnels (Distance Vector Multicast Routing
  Protocol)
• IGMP (Internet Group Management Protocol)
• Broadcast Prune model
• Source rooted distribution tree
• Problems scalability, stability, QoS (rate
  limited tunnels), Virtual topology (tunnel)
  management.

6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
Intra-domain Multicast routing

• Sparse mode protocols
• MOSPF
• PIM-SM
• CBT
• Based on explicit join, unlike dense mode
  protocols (broadcast prune)
• shared distribution tree (uses a core or
  Rendez-vous Point (RP))

20
Inter-domain Multicast routing (1)

• Carrying IP Multicast routes in BGP (MBGP),
• straightforward extension of BGP, also called
  BGP (RFC2283)
• Subsequent Address Family Identifier (SAFI) added
  to BGPs reachable/unreachable messages.
• SAFI can specify unicast, multicast or
  unicast/multicast forwarding information.

21
Inter-domain Multicast routing (2)

• Multicast Source Discovery Protocol (MSDP)
• allows to connect domains running multicast
  sparse mode protocols in another way than sharing
  a common RP.
• Runs off the RP router, similar to BGP, MDSP
  flooding of SA (Source Active) messages (peer-RPF
  flooding)
• problems latency (propagation of SA messages),
  scalability

22
Inter-domain Multicast routing (3)

• MBGP/MDSP well introduced in vBNS/Abilene and
  (part of)TEN-155.
• Mbone (AS10888) interworking via NASA Ames
  Multicast Friendly Internet Exchange (MIX)
• the MIX provides connectivity between the Mbone
  and 10 other ASs who peer using PIM-SM/MBGP/MSDP

23
Longer term Multicast Proposals

• Border Gateway Multicast Protocol (BGMP)
• bi-directional shared trees between domains
• MASC (Multicast Address-Set Claim) to allocate
  Multicast group addresses in such a way as to
  avoid collisions.
• GLOP
• Root Addressed Multicast Architecture (RAMA)
• Express Multicast
• Simple Multicast
• Single Source Multicast (SSM)
• Removes the requirements of having IGMP capable
  hosts
• Better scalability (source based dist. tree)
• Small Group Multicast (SGM)

24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
Why is it a difficult problem?

• IPv4 is a huge success story 150 millions
  hosts
• IPv6 "had" a chicken and egg problem.
• Early IPv6 adopters face Meltcafe law
• "The value of a network is proportionalto the
  square of the number of users."
• We are still very early in the transition.
• We are talking about a 10 year process

34
Incremental deployment on existing infrastructure

• Do not disrupt IPv4 traffic
• Use IPv4 networks as of today
• Use IPv6 to re-establishend-to-end IP
  connectivitygt The rest of this presentation
  assumes this strategy

35
Two set of problems

• Communications between IPv6 islands
• Dual stack routers
• and
• IPv6 over IPv4 tunnels
• Communications between IPv6 and Ipv4 hosts
• Dual stacks,
• Application Level Gateways (ALG)
• Network Address Translators (NAT)
• Temporary allocation of IPv4 Addresses IPv4 in
  IPv6 tunneling

36
Basic mechanismHybrid stack (a.k.a. dual stack)

• A node can "talk" IPv4 IPv6.
• New applications support both IPv4 IPv6.
• No need for two set of applications, one for v4,
  one for v6.

Application
TCP/UDP
IPv6
IPv4
LAN
37
Basic mechanismTunnel

• IPv6 packets are "encapsulated" withinIPv4
  packets.

TCP/UDP
Payload
IPv4
IPv6
38
Connecting two IPv6 clouds Configured tunnel
IPv4
tunnel
IPv6
IPv6
hybrid stack routers
39
Connecting an isolated hostTunnel Broker/1
tunnel broker
IPv4 Internet
tunnel request
IPv6 Internet
isolated hybrid stack host
40
Connecting an isolated hostTunnel Broker/2
tunnel broker
IPv4 Internet
tunnel config
tunnel config
IPv6 Internet
isolated hybrid stack host
tunnel server
41
Connecting an isolated hostTunnel Broker/3
IPv4 Internet
IPv6 Internet
tunnel
isolated hybrid stack host
tunnel server
42
6to4 mechanism

• One IPv4 global address gt one /48 IPv6
  site prefix

ISPv4 assigned
managed
auto-configured
pre-defined
2002
IPv4
SLA
Interface ID
48 bits
16 bits
64 bits

• Stateless tunnels span over the IPv4
  infrastructure without configuration to reach
  other 6to4 domains.

43
Building automatic extranets Connecting IPv6
clouds with 6to4
6to4 hybrid- stack routers
IPv4
6to4
tunnels
44
What are Content Delivery Networks

• A centrally managed overlay network of devices
  that collectively facilitate the delivery of
  content to end users
• Solve network bandwidth bottleneck
• Solve server throughput bottleneck
• In principle, greatly improves the overall
  response time, also makes it possible to deliver
  good quality real time streams under almost any
  conditions!
• An isolated Content Engine can be seen as a
  simple instance of a CDN.

45
(No Transcript)
46
Technology Components

• Content distribution
• Placing the content to the devices
• Request routing
• Steer users to a delivery node that is close
• Content delivery
• Protocol processing, access control, QoS
  mechanisms
• Resource accounting
• Logging and billing

47
Content Distribution

• Goal
• position content objects into delivery devices
• Different content types use different techniques
• Static images and texts pulled cached, or
  pushed
• Multimedia contents usually pre-positioned
• Dynamic pages requires prior setup

48
Distribution Mechanisms

• HTTP request for pulling
• Example standard HTTP reverse proxy
• FTP of tar files
• Some equipment vendors use this technique
• Rate limited tree-form replication
• Example Ciscos Soda algorithm

49
Distribution Mechanisms using Multicast

• Application-level reliable multicast
• Example Inktomis Fast-Forward
• Unreliable IP multicast with file-level error
  correction
• Example Digital Fountain, multicast-ftp
• Unreliable IP multicast
• Example RealNetworks

50
Content Consistency Mechanisms

• Expiration times or TTL
• Renaming in the HTML file
• Web Cache Invalidation Protocol (WCIP)
• Nodes receive invalidations when objects change
• Objects are organized into channels
• Nodes subscribe to a channel to receive
  invalidation

51
Request Routing

• Goal steer the client such that it fetches the
  content from a close node
• Methods
• DNS selection
• HTTP redirection
• Transparent interception

52
Overview of Request Arrival Process
How a request for www.xyz.com/index.html arrives
at 1.2.3.4
Root NS
2. where is name server of xyz.com?
DNS server
3. NS record 1.2.3.1
4. what is IP of www.xyz.com?
1. what is IP addr of www.xyz.com?
xyz.com NS
IP 1.2.3.1
5. A record1.2.3.4
6. 1.2.3.4
Client
s w i t c h
7. GET /index.html
Router
Server
IP 1.2.3.4
53
DNS selection

• Basic idea xyz.coms NS returns node close to
  client
• How to become xyz.coms NS?
• Rewrite URLs (aka Akamizer)
• Take a subdomain cdn.xyz.com and put all content
  there
• Accuracy limited to clients name server
• Only suitable for ISP or overlay networks
• Not suitable for some enterprise or cable networks

54
HTTP Redirection

• Basic idea web server tells client to go
  somewhere else
• Returns 302 redirect 1.2.4.5/index.html
• Mostly used for multimedia objects
• These objects are usually put together in an
  index file (.sml or .asx) and clients fetch the
  index file via HTTP before streaming
• Accuracy is at individual client level
• More suitable for enterprise and cable networks

55
Transparent Interception

• Router and switch along the request path can send
  the request elsewhere
• Mostly used for distributed data centers
  front-ended with L7 switches
• Example Ciscos CSS11k WebNS

56
Algorithms for Request Routing

• Map-based
• Create a map of the Internet based on AS domains,
  pick the node with the shortest hop count to
  client
• Or, set up coverage zones mapping a node to a
  collection of subnets
• Racing-based
• Let the delivery nodes all race to the client
  with A-records
• Winner is selected by client automatically

57
The Boomerang Algorithm

• Ciscos research published in WCW01
• xyz.coms NS server forwards lookup of
  www.xyz.com to all delivery nodes
• Delivery nodes all send A record response with
  its own IP address to the client
• The one that reaches the client first wins
• NS server times the forwarding so that lookup
  message arrives at all nodes around the same time

58
Interaction between Content Distribution and
Request Routing

• Dont route request to a node that doesnt have
  the content!
• Particularly important for large streaming
  contents
• Such content are usually pre-positioned to ensure
  high-bandwidth playbacks
• Nodes need to report its content acquisition
  status to the request router

59
Content Delivery

• Goal serve content to each client at desired
  quality of service
• Supported protocols
• HTTP
• Microsoft MMS
• Open standard RTP/RTSP
• RealNetworks RTP/RTSP
• Usually part of the larger CDN system

60
Content Access Control

• Content object attributes
• Publication date and Expiration date
• ACL based on user/group/IP
• User authentication
• HTTP basic
• Microsoft NTLM for enterprise environment
• other schemes
• Media Rights Management

61
QoS of Content Delivery

• Server QoS
• Server needs to make sure it has enough CPU and
  disk to service the stream at specified bit rate
• Network QoS
• Interoperate with routers via DiffServ bits
• Coordination with request router
• delivery devices should communicate load
  information to the Request Router

62
Summary

• Main components of building a CDN
• Content distribution
• Request routing
• Content Delivery
• Resource accounting
• A CDN system requires the four components to work
  in concert with each other!
• Cisco is the only vendor that provide the full
  solution!

63
WCCP Overview

• Transparent interception of selected protocols
• e.g. Web (port 80), or RealAudio/Video
• Subject to access lists restrictions
• Scalable versatile solution (transparent
  cache/reverse proxy)
• Cache Engine selection on hashed destination
  address (normal Web caching mode)
• Or hashed source address (reverse proxy mode)
• Also called Web acceleration
• Hierarchical deployment
• Other features
• Overload bypass
• Dynamic client bypass (IP authenticated requests)
• WCCP licensed to many companies
• WCCPv1 is an IETF draft, not sure about WCCPv2

64
Cache Engine vs Content Engine(a personal view)

• Transparent caches are a big improvement over
  conventional caches and traditional cache
  hierarchies using ICP (Internet Cache Protocol),
  however.
• Content engines are a natural evolution of
  conventional Web caches also reflecting the
  evolution of the Internet towards real time
  services (streaming audio video, video on
  demand, etc).
• Ciscos Content Engine incorporates a Cache
  Engine as well as a Real Networks
  server/splitter.

65
Cisco Cache Engine
66
(No Transcript)
67
(No Transcript)
68
(No Transcript)
69
(No Transcript)
70
(No Transcript)
71
(No Transcript)
72
(No Transcript)