500 likes | 637 Views
Updates of Study Progress. Othman Othman M.M. 30/1/2012. Outline:. Advisory Committee Meeting: Content Anycasting. OpenFlow Enhancement. Current Status . Next Research. Outline:. Advisory Committee Meeting: Content Anycasting. OpenFlow Enhancement. Current Status . Next Research.
E N D
Updates of Study Progress Othman Othman M.M. 30/1/2012
Outline: • Advisory Committee Meeting: • Content Anycasting. • OpenFlow Enhancement. • Current Status. • Next Research.
Outline: • Advisory Committee Meeting: • Content Anycasting. • OpenFlow Enhancement. • Current Status. • Next Research.
1-Content Anycasting: • Outline: • Goal . • Current Technologies. • Advantages of Combining Technologies. • How to Combine Technologies. • Current Technologies’ Scenarios. • Content Anycast Scenario. • Evaluation. • Anycast Comparison. • P2P Comparison. • Conclusion.
1-Content Anycasting:1-Goal: • To have contents with high availability. • Improving the availabilityof the content server. • To improve the overall usage of bandwidth of the whole network. • Nowadays many Future Internet researches, technologies going. • OpenFlow is one of candidate Future Internet technologies. • to create Circuit based like systems. • To support mobility, computing centers ……
1-Content Anycasting:2-Current Technologies. • Anycast: • Multiple nodes with the same address (Sa). • Packet sent to (Sa) will be delivered to the node with nearest location. • Peer to Peer: • Depends on user clients to provide service. • Implements an application layer overlaynetwork. Sa Sa Sa Fig 1. Anycast Fig 2. BitTorrent Source :http://en.wikipedia.org/wiki/Image:Torrentcomp_small.gif
1-Content Anycasting:2-Current Technologies. • Peer to Peer: • Overhead because of overlay nature, protocol, peer discovery, and looking up in index. • Anycast: • All of the content servers must have identical contents. • Lacks the flexibility, and not dynamic. 10.10.10.1 192.168.0.1 Cont 1 Cont 1 Cont 2 Cont 2 Router 1 Cont 3 Cont 3 Router 2 Fig 2. BitTorrent Source : http://computer.howstuffworks.com/bittorrent2.htm Cont n Cont n 10.10.10.1 Router 3 Router 4 192.168.0.2 Routing Table : Destination Next-Hop Distance 192.168.0.0 127.0.0.1 0 10.0.0.1 192.168.0.1 1 10.0.0.1 192.168.0.2 2
1-Content Anycasting:3-Advantages of Combining Technologies. • Anycast. • Content Centric Networks. • Peer to Peer. • Combination of those technologies: • Users contributing in service: to improve the availability, and improves the overall use of bandwidth in whole network. • Choosing destination by network:to remove the burden of finding destination and thus a faster response. • Content ID: to have more flexibility down to the level of contents rather than the node level.
1-Content Anycasting:4-How to Combine Technologies. • Content anycasting does its role by using: • OpenFlow for the process of choosing the destination, along with the aid of the content server. • A new procedure for requesting content is introduced to enable the clients to get the contents. AnycastManager Content Server Su Client A Client B OpenFlow Router Au Bu
1-Content Anycasting:4-How to Combine Technologies. • The new procedure for getting the content: • Phase 1: getting the content ID. (e.g.: from the URL). • Phase2: using the content ID in Probe protocol. • Phase 3: getting file via TCP. OpenFlow router New Client Current Client Destination: serverIP :START Phase 2 Destination: CurrentClientIP :START START/ACK With CurrentClientIP in the probe header ACK / ACK Destination: CurrentClientIP Phase 3 TCP Session Fig 1: Procedure for getting content (Hand shake of the Probe Protocol)
1-Content Anycasting:5-Current Technologies’ Scenarios. • Anycast: Client will send packet to 10.10.10..1 10.10.10.1 10.10.10.1 Cont 1 Cont 1 Cont 1 Cont 2 Cont 2 Cont 2 Cont 3 Cont 3 Cont 3 Cont n Cont n Cont n Network will choose end node based on routing measures 10.10.10.1
1-Content Anycasting:5-Current Technologies’ Scenarios. • P2P; BitTorrent: Tracker 100% Seeder 70% 30% 100% Seeder 0%
1-Content Anycasting:6-Content Anycast Scenario. Threshold reached Redirection Req. AnycastManager Content Server Redirection To: Au Cont id=X Content ID = X, in URL Su Get Content ID form URL To: Su Cont id=X Client A TCP Session Client B Au Bu OpenFlow Router
1-Content Anycasting:7-1Evaluation: Anycast Comparison. • Using simple simulator, built using Java. • Simulates a network with 5 areas (5 networks); • Content Anycast: use only one server. • Regular Anycast: 5 replica servers each in one network. • Measure the server load : number of connections the server serves. • Shows that : • Regular Client server: single server load is 100%(one server for all) • Regular anycast: each server out of 5 load is 20%. • Content Anycast : single server load is 50%,33%, 25% and 20%respectively for case of client can serve 1, 2, 3, 4 other clients.
1-Content Anycasting:7-1Evaluation: Anycast Comparison. • Content Anycast can achieve same load as regular anycast but using only one server instead of 5.
1-Content Anycasting:7-2Evaluation: P2P Comparison. • Content Anycast, average hops = 5. • Regular P2P, average hops = 14. • Content Anycast has a lower hop count due to using pre-installed redirections that lead request to peer in same network. • To evaluate start-up time: • Count the number of hops that the client request travels across the network + hops reply travels.
1-Content Anycasting:7-2Evaluation: P2P Comparison. Server load limit • To evaluate serve/management entity load. • Generate flash crowds periodically every 100 cycle. • Count the number of peer quivery requests received by the server/ management entity. • Content Anycast has a lower load due to using pre-installed redirections that lead request to peer in same network rather than the server.
1-Content Anycasting:8-Conclusion. • New mechanism for requesting content is designed to enable content anycasting. • Using the content id in the process of getting the content. • A Probe protocol is designed (modification to UDP) to be used. • Simulation shows that Content anycast can • reduce load to match reduction of the regular anycast (under some conditions) using only one server. • On simulation scenario: 80% reduction in number of servers. • Achieve shorter start-up time and less load on the management entity/ server compared to P2P. • On simulation scenario: 74% reduction in number of hops needed to get contents.
Outline: • Advisory Committee Meeting: • Content Anycasting. • OpenFlow Enhancement. • Current Status. • Next Research.
2-OpenFlow Enhancement: • Outline: • Motivation and Goal. • An attempt to solve the problem (3 enhancements). • First: Network Equipment to Equipment flow installation. • Second: Low Level Header Description. • Third: Inactive Flows. • Flow Aggregation Algorithm. • Evaluation. • Conclusion.
2-OpenFlow Enhancement:1-Motivation and Goal. • Tight coupling between OpenFlow switch and controller. • Every thing is up to the controller. • Controller might be bottleneck. • number of flows that can be installed by the NOX controller as shown in [1] are 30K flow/sec, and the flow arrival rate in [2] that is 100K flow per second. • OpenFlow supports MAC, VLAN, IP, TCP, UDP. • Better to have; more room for researchers. [1].Tavakoli, A., Casado, M., Koponen, T., & Shenker, S. (n.d.). Applying NOX to the Datacenter. Proc. HotNets (October 2009). [2]. Kandula, S., Sengupta, S., Greenberg, A., Patel, P., & Chaiken, R. (2009). The nature of data center traffic: measurements & analysis. Proceedings of the 9th ACM SIGCOMM conference on Internet measurement conference (p. 202–208). ACM.
2-OpenFlow Enhancement:1-Motivation and Goal. • Improve OpenFlow. • Support self-reactive behavior. • Reduce load on controller. • Giving researchers more flexibility. • Step towards having wider adoption of OpenFlow. • If OpenFlow is thought of as one of the Future Internet technologies, However some debate that OpenFlow have some limitations. • So our aim is to enhance OpenFlow, to make it more suitable for wider adoption and implementation within networks or in whole OpenFlow networks.
2-OpenFlow Enhancement:2-An attempt to solve the problem. • Network equipment to Network equipment Flow Programming: • To create traffic-aware self-reactive network. • Can be used to delegate some flows to less loaded network equipment. • To easily program whole network without loading controller. • Low level description of Headers: • To easily adopt new protocols. • To be able to program flows in very exact way. • New type of Flows: • Programed as inactive flows and later activated by the flow. • To provide the controller with a more relaxed way to handle precisely timed tasks. • Can cooperate with Device to Device programming.
2-OpenFlow Enhancement:2-1 First: Network Equipment to Equipment flow installation. Flows to manipulate headers in packets P P Packet P Packet P P P Packet Packet Packet Packet • To reduce load off the controller. • Give the equipment ability to act by their own to reduce load off loaded equipment. • Alternative way to install flows to whole network (e-e propagation). Fig1. Equipment overloaded, due to many flows to carry out. Packet P Packet PE Packet PE PE P P Fig2. Overloaded equipment delegates some flows to other equipment. Flows to manipulate headers in packets Fig3. Reduced load off the overloaded equipment.
2-OpenFlow Enhancement:2-2 Second: Low Level Header Description. • OpenFlow can deal with headers of: • Ethernet, IP, TCP, UDP, ICMP, ARP, VLAN • This limits the usage to those protocols. • Using Low level definition of header fields can be more usable in case of new protocols. Fig2: Low level description of headers Fig1: Fields from packets used to match against flow entries As shown in : OpenFlow Switch Specification Version 1.0.0, December 31, 2009
2-OpenFlow Enhancement:2-3 Third: Inactive Flows. • Original OpenFlow: flows activated by default, controller keeps track of time. • Initially installed as inactive. (not usable). • Activated on right time, by: • Explicit activation packet. • Activation Flow. • Preset time. Fig1. Migration and Redirection using OpenFlow. Fig2. Delay due to controller overload in Migration. Inactive Flows Fig3. Migration and Redirection using OpenFlow and Inactive Flows. Migration Migration Migration Migration Flows activation Migration Migration ?
2-OpenFlow Enhancement:3-Flow Aggregation Algorithm. Flow Table • How to delegate flows? • Aggregate flows that have common features, and responsible for some portion of traffic. • i.e. to aggregate many flows to one. • Delegate the aggregated flows to other equipment. • Use Flow Aggregation Algorithm. • Overloaded equipment flows = original flows – delegated flows. Flow Aggregation Algorithm aggregated flow (one or more)
2-OpenFlow Enhancement:3-Flow Aggregation Algorithm. Start Build Histograms for all Fields None Strict Aggregation percentage? Wide Aggregate SrcIP Strict None Wide Aggregate DstIP Wide Strict Find commonvalues from two wide aggregations. None, Wide Strict None Finish Fail
2-OpenFlow Enhancement:4-Evaluation. • Java Program to evaluate the efficiency of Flow Aggregation Algorithm. • FAA success rate of aggregation = 79.7 % • NS3 simulation to evaluate: • Controller load reduction . • Efficiency in reducing load off overloaded equipment. • Traffic generated due to the new enhancements. • Compare the enhanced OpenFlow with the current OpenFlow by comparing; controller load, OpenFlow equipment load.
2-OpenFlow Enhancement:5-Conclusion. • Aim to improve OpenFlow by reducing load off the controller, make it self-aware and self-reactive, enable researchers to support their own protocols, and enable an easier support for time critical functions. • Achieving goals by proposing 3 new enhancements to OpenFlow: • Network equipment to equipment flow installation. • Low level Header description. • Inactive flows. • Proposing Flow Aggregation Algorithm, to enable the enhancements. • Simulation shows the success rate of FAA is 79.7 %
Outline: • Advisory Committee Meeting: • Content Anycasting. • OpenFlow Enhancement. • Current Status. • Next Research.
3-Current Status: • Content Anycasting. • IEICE. • Title of Special Section: Architectures, Protocols, and Applications for the Future Internet. • Title: On Demand Content Anycasting to Enhance Content Server Using P2P Network. • Volume and Number: Vol.E95-D,No.2,pp.-,Feb. 2012. • OpenFlow Enhancement. • Working on the evaluation. • Next Research. • Working on design.
Outline: • Advisory Committee Meeting: • Content Anycasting. • OpenFlow Enhancement. • Current Status. • Next Research.
4- Next Research: • Proposing a design for Future Internet. • Current internet faces many difficulties: • Mobility. • Multi-homing. • Multicasting and Anycasting. • Routing table growth. • Security. • Accountability. • A totally new design (clean slate) ; would be easier to solve the problems than add-on .
4- Next Research: • An example scenario that shows the motivation. • Current Internet. • Target Future Internet. Traffic Transfer agreement Traffic Transfer agreement VPN Connection Connection Traffic Traffic of the mobile node will be treated same as it would be when the node was in its original network.
4- Next Research: • Initial Design Steps: • Gathering information about current internet’s problems. • Making expectations for the Future Internet. (How we expect it to be, what we want it to handle). • Designing the Naming and addressing and their spaces. • Designing rules or guidelines that governs the naming and addressing. • Designing data transfer methods and mechanisms (such as; routing, delivering data, ….).
Outline: • Advisory Committee Meeting: • Content Anycasting. • OpenFlow Enhancement. • Current Status. • Next Research.
1-Content Anycasting: • Outline: • Goal . • Anycast vs. Multicast. • Current Technologies. • Advantages of Combining Technologies. • How to Combine Technologies. • Current Technologies’ Scenarios. • Content Anycast Scenario. • Evaluation. • Anycast Comparison. • P2P Comparison. • Conclusion.
1-Content Anycasting:3-OpenFlow Overview. • Separates routing decision making (in controller) and the forwarding (in the switch or router). • Matching in the switch or router is done according to Layer 2, 3 and VLAN headers. Figure 1: OpenFlow switch (ref: Nick McKeown et al, “OpenFlow: enabling innovation in campus networks”)
1-Content Anycasting:2-Anycast vs. Multicast. • Multicast: • Deliver multiple copies of packetsto multiple hosts. • Anycast: • Deliver original packetto oneout a groupof hosts. • Deliver to nearest node depending on routing measures. Multicast Anycast
1-Content Anycasting:5-Advantages of Combining Technologies. • Anycast. • Content Centric Networks. • Peer to Peer. • Combination of those technologies: • Users contributing in service: to improve the availability, and improves the overall use of bandwidth in whole network. User contribution In Content Anycasting Regular Client/Server
1-Content Anycasting:5-Advantages of Combining Technologies. • Anycast. • Content Centric Networks. • Peer to Peer. • Combination of those technologies: • Choosing destination by network: to remove the burden of finding destination and thus a faster response. Tracker ? Content Server OpenFlow Router Choosing Destination By Network In Content Anycasting Tracker Role in BitTorrent
1-Content Anycasting:5-Advantages of Combining Technologies. • Anycast. • Content Centric Networks. • Peer to Peer. • Combination of those technologies: • Content ID:to have more flexibility down to the level of contents rather than the node level. • Decision based on: • IP address. • sometimes port# • Decision based on: • IP address. • Content ID. Content Server Content Server OpenFlow Router Using Content ID In Content Anycasting Regular
4-2- Inputs of the FAA. Flow Table • Input is Flow Table. • Flow table entry have one or more fields form Fig 1. • Src IP and Dst IP can be wildcarded as defined by OpenFlow Flow Aggregation Algorithm aggregated flow (one or more) Fig1: Fields from packets used to match against flow entries As shown in : OpenFlow Switch Specification Version 1.0.0, December 31, 2009
4-3-Outputs of the FAA. Flow Table • Aggregated Flow can use: • Parts of Src IP, Dst IP using wildcards. • Any other field as an exact match. Range of portions of total traffic e.g. (20%-30%) Flow Aggregation Algorithm aggregated flow (one or more) Fig1: Fields from packets used to match against flow entries As shown in : OpenFlow Switch Specification Version 1.0.0, December 31, 2009
2-OpenFlow Enhancement:4-Flow Aggregation Algorithm. Start Start 1 Build Histograms for all Fields Add the most significant bit of each IP in each flow to trie. If exist increment trie node counter. Else create node and increment counter • By Building a Trie. • Find length of wildcard. • Value of wildcarded IP. None Strict Aggregation %? 2 Wide Aggregate SrcIP Is aggregation successful? Yes Strict None 3 Wide Aggregate DstIP No Add the next less significant bit of each IP in each flow to trie. If exist increment trie node counter. Else create node and increment counter Is memory limit reached? None 4 No Strict 5 Wide Find commonvalues from two wide aggregations. Yes Mark the trie node that has the best ratio . 6 None, Wide Finish 7 Strict Finish Fail
2-OpenFlow Enhancement:4-Flow Aggregation Algorithm. Start Build Histograms for all Fields Start • By building Histogram. 1 None Strict Aggregation %? Build Histogram 2 Wide Aggregate SrcIP Is aggregation successful? Strict None 3 Wide No Aggregate DstIP Yes Find nodes with right aggregation ratio None 4 Strict Wide Finish 7 Find commonvalues from two wide aggregations. None, Wide Strict Finish Fail