distributed dynamic replica placement and request redirection in content delivery networks advisor :...

23
Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Upload: abigayle-boone

Post on 13-Dec-2015

221 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks

Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks

Advisor : Ho-Ting Wu

Student : Yu-Chiang Lin

Date:2011/5/30

Page 2: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 2

OUTLINE CDN introduce

Distributed Dynamic Replica Placement

Cloning of a replica

Replica removal

Redirection

Future work

Reference

Page 3: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 3

OUTLINE CDN introduce

Distributed Dynamic Replica Placement

Cloning of a replica

Replica removal

Redirection

Future work

Reference

Page 4: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 4

CDN Introduce

Fig:Server Farm(source from EBU technical review)

Page 5: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 5

CDN Introduce

Consider TCP transmittion, throughput may affect by latency or packet lost。

Page 6: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 6

CDN Introduce The Content Delivery Networks (CDN) paradigm is

based on the idea to transparently move third-party content closer to the users

content is replicated on CDN servers which are located close to the final users so that users perceive a better content access service.

Page 7: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 7

CDN Introduce

Fig:Content Delivery Network(source from EBU technical review)

Page 8: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 8

CDN Introduce

Fig:Content Delivery Network

Page 9: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 9

CDN Introduce

Four Important technique

1.Content route

2.Content distribution

3.Content store

4.Content management

Page 10: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 10

CDN Introduce CDN issue

1) Deciding the kind of content that should be hosted (replica placement)

2) selecting the best replica for a given user

3) designing mechanisms for transparent redirection of the users requests to the best replicas

Page 11: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 11

OUTLINE CDN introduce

Distributed Dynamic Replica Placement

Cloning of a replica

Replica removal

Redirection

Future work

Reference

Page 12: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 12

Distributed Dynamic Replica

Distributed scheme to allocate and deallocate replicas, so that the user requests are satisfied while minimizing the CDN costs in a dynamic scenario

This scheme always accounts for the current replica placement,adding replicas or changing replica location only when needed.

Each site j V∈ R autonomously decides on whether some of the replicas it stores should be cloned or removed.

Page 13: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 13

Problem Formulation VA: access nodes

VR: CDN servers sites

d(i, j): user (access node) i to a replica j

dmax: distance threshold

xi,c: the volume of user requests originated at node i V∈ A for content c C∈

Umax, Umid, Ulow: load threshold

Page 14: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 14

Cloning of a replica Function add_replica(j, c)

1: lj,c =i V∈ A αij,c · xi,c

2: while lj,c/rj,c− Umax > 0

3: best_served = 0

4: best_distance = ∞

5: best_vr = undefined

6: for all j’ ρ∈ (j) s.t. rj,c < VmaxR do

Page 15: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 15

Cloning of a replica 7: l’j,c =i α∈ (j’) αij,c · xi,c

8: total_distance =i α∈ (j’)αij,c · xi,c · di,j’

9: if (l’j’,c < best_served) ∨

10: (l’j’,c = best_served ∧

11: total_distance < best_distance) then

12: best_distance = total_distance

13: best_served = l’j’,c

Page 16: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 16

Cloning of a replica 14: best_vr = j’

15: end if

16: end for

17: if best_vr = undefined then

18: exit

19: end if

20: ask best_vr to add a replica

Page 17: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 17

Cloning of a replica 21: compute l’’bestvr,c = min(i α∈ (bestvr) αij,c · xi,c, 1)

22: lj,c = lj,c − l’’bestvr,c

23: remove from the set of requests those that can be offloaded

24: end while

Page 18: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 18

Replica removal A replica can be removed if (and only if) it serves no

requests

A replica is removed only when it has not been serving requests for a time long enough to bring the exponential average down to zero.

Page 19: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 19

Redirection Based on this feedback users requests are directed

away from a replica in response to threshold events (if the replica load exceeds Umax or falls below Ulow)

As an example, an underloaded replica informs the redirection system which then tries to offload requests to some other replicas (if possible)

A perfect load balancing may be impossible due to the distance constraint

Page 20: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 20

Redirection

Fig. A model for the redirection scheme

Page 21: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 21

OUTLINECDN introduce

Distributed Dynamic Replica Placement Cloning of a replica

Replica removal

Redirection

Future work

Reference

Page 22: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 22

Future Work comparison of different settings of the Umid

observe dmax and add remove replica relationship

Page 23: Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks Advisor : Ho-Ting Wu Student : Yu-Chiang Lin Date:2011/5/30

Page 23

Reference [1] N. Bartolini, F. Lo Presti, and C. Petrioli, “Optimal dynamic r

eplica placement in Content Delivery Networks,” in Proceedings of the 11th IEEE International Conference on Networks, ICON 2003, Sydney, Australia, September 28–October 1 2003, pp. 125–130.

[2] F. Lo Presti, C. Petrioli, and C. Vicari, “Dynamic replica placement in content delivery networks,” in Proceedings of MASCOTS 05, September 2005.

[3] F. Lo Presti, C. Petrioli, and C. Vicari, “Distributed Dynamic Replica Placement and Request Redirection in Content Delivery Networks ,” in Proceedings of MASCOTS 07, 2008.