Fig 2 - uploaded by Lixia Zhang
Content may be subject to copyright.
Multicasting requests with a limited hop count reduce the effectiveness of request suppression. (Scenario in the star topology).
Source publication
Scalable reliable multicast (SRM) is a framework for reliable
multicast delivery. In order to maximize the collaboration among the
group members in error recovery, both retransmission requests and
replies are multicast to the entire group. While SRM effectively uses
random timers to suppress duplicate requests and replies, the global
nature of the...
Contexts in source publication
Context 1
... scenario illustrated in Fig. 2 can cause much higher request overhead. All the requesters in Fig. 2 choose as the replier. They multicast their requests with a hop count value that is big enough to reach but not big enough to reach one another for duplicate suppression. In the worst case, the number of requests per loss within the loss region is equal to the number ...
Context 2
... scenario illustrated in Fig. 2 can cause much higher request overhead. All the requesters in Fig. 2 choose as the replier. They multicast their requests with a hop count value that is big enough to reach but not big enough to reach one another for duplicate suppression. In the worst case, the number of requests per loss within the loss region is equal to the number of requesters. Therefore, to minimize the number of duplicate ...
Context 3
... In the worst case, the number of requests per loss within the loss region is equal to the number of requesters. Therefore, to minimize the number of duplicate requests, requesters have to set their request hop counts large enough not only to reach the replier but also to reach one another for suppression. 5 The characteristic of this scenario in Fig. 2 is that multiple requesters do not choose one another as repliers for retrans- mission. For example, both and request who is closer to the source than and . Because the replier selection is autonomous and independent, in general a member will not know the selection of other members. However, two members and are very likely to be in the ...
Context 4
... requesters do not choose one another as repliers for retrans- mission. For example, both and request who is closer to the source than and . Because the replier selection is autonomous and independent, in general a member will not know the selection of other members. However, two members and are very likely to be in the scenario described in Fig. 2 if they fall within the following three conditions: 1) requests someone closer to the source than ; 2) requests someone closer to the source than ...
Context 5
... large request scopes in both the hop- scoped and the group-scoped error recoveries because their requests have to propagate across the Mbone to reach helpers. On the other hand, has relatively small request scopes because its requests only have to reach . have relatively large requests scopes because they are involved in the scenario sketched in Fig. 2. They have to extend their request hop counts to suppress one ...
Citations
... As a result, data is multicast to a predefined group of nodes, and this constraint makes it difficult to achieve efficient data dissemination to a group of nodes with different resource constraints. This mismatch also makes efficient loss recovery difficult: reliable multicast desires re-transmissions of specific ADUs by nearby members [25], while routers have no concept of ADUs and deliver all packets to the whole group. ...
Distributed dataset synchronization, or Sync in short, plays the role of a transport service in the Named Data Networking (NDN) architecture. A number of NDN Sync protocols have been developed over the last decade. In this paper, we conduct a systematic examination of NDN Sync protocol designs, identify common design patterns, reveal insights behind different design approaches, and collect lessons learned over the years. We show that (i) each Sync protocol can be characterized by its design decisions on three basic components - dataset namespace representation, namespace encoding for sharing, and change notification mechanism, and (ii) two or three types of choices have been observed for each design component. Through analysis and experimental evaluation, we reveal how different design choices influence the latency, reliability, overhead, and security of dataset synchronization. We also discuss the relationship between transport and application naming, the implications of namespace encoding for Sync group scalability, and the fundamental reason behind the need for Sync Interest multicast.
... Switching between unicast and IP-multicast retransmission based on the number of requests has also been considered . [ [22] recovery if the topology of a session is star-shaped, and the group-scoped error recovery has more overhead on the members as well as on the underlying IP-multicast routing, as it requires individual members to maintain and manage multiple local IP-multicast groups. Complexity and practical deployment is again a major obstacle. ...
IP-multicast is a bandwidth efficient transmission mechanism for group communications. Reliability in IP-multicast, however, poses a set of significant challenges. To address the reliability and scalability issues in IP-multicast, this paper proposes a novel, highly distributed, and lightweight overlay peer-to-peer retransmission architecture that exploits path-diversity by taking advantages of both IP-multicast and an overlay network. An approach that leverages both disjoint-path-finding and periodic selective probing to take into account peer's recent packet loss probability, retransmission delay and recent retransmission success rate is proposed to effectively construct an efficient and dynamic overlay retransmission network. We show that the proposed path diversity overlay retransmission architecture has the potential to significantly reduce the retransmission delay, improve the reliability, playback quality, and scalability of IP-multicast based multimedia applications. Given a deployed IP-multicast network, the proposed overlay retransmission architecture is practical, scalable, and easy to deploy, requiring no change to the existing network infrastructure.
... The main advantage of the scheme is that bandwidth consumption for retransmission can be bound to a constant times of the lower bound bandwidth. We have formally proved that boundedness of BestRelay retransmission tree at Sect. 2. SRM [3], [9], [10] takes multicast-based reliable multicast for many-to-many multicast environment. When a host detects a packet loss, it multicast NAK to entire group of receivers. ...
In this paper, we address how to construct efficient retransmission trees for reliable multicast. Efficiency of retransmission trees mainly depends on locations of repairers, which are in charge of retransmitting lost packets. We propose an algorithm for each receiver to find a repairer for efficient recovery. The resulting tree for retransmission is organized by pairs of a receiver and a repairer which is the host "nearest" to the reciever among the multicast group members "nearer" to the sender. We formally prove that the proposed algorithm realizes reliable multicast with only constant times of a lower bound cost achievable through impractical router support. We also evaluate the algorithm through extensive simulations.
... For example, when a new member in a different geographic region joins the session and K G groups currently exist, we let the member increase K G by one and create a new group without merging two groups who are located nearest but have different path characteristics. When the distribution of session members are very sparse, the ACK tree structure of GAM becomes close to the source specific tree structure of SRM for local error recovery [17]. In this case, reduction of the tree maintenance overhead might not be so substantial. ...
... The timer-based suppression might not work so much well in WAN or for large number of participants. Local recovery based TTL scoping is suggested [17] . For stable delivery, SRM's implementation requires that every node should store all packets or that the application layer should store all relevant data. ...
In this paper, we present the design, implementation, and performance analysis of Group-Aided Multicast (GAM), a scalable many-to-many reliable multicast transport protocol. It achieves high quality ACK trees while keeping the tree maintenance overhead reasonably low in the presence of dynamic group membership and route change. The proposed scheme is supported by a group configuration mechanism that organizes the members in a multicast session into multiple small groups and a tree configuration mechanism that maintains the logical trees according to the underlying multicast routing trees. With the two mechanisms, it builds a two-layer hierarchy of multi-level logical trees from which high-quality per-source ACK trees are generated. Simulation results show that the proposed protocol is more scalable than an existing protocol in terms of processing time for request/repair messages and recovery latency.
... However, lack of scoping means that requests and retransmissions generated by SRM will reach the entire group. Local recovery methods have been proposed for SRM [20], which bring SRM closer to our ALH scheme. ...
One approach to achieving scalability in reliable multicast is to use a hierarchy. A hierarchy can be established at the application level, or by using router-assist. With router-assist we have more fine-grain control over the placement of error-recovery functionality, therefore, a hierarchy produced by assistance from the routers is expected to have better performance. In this paper, we test this hypothesis by comparing two schemes, one that uses an application-level hierarchy (ALH) and another that uses router-assisted hierarchy (RAH). Contrary to our expectations, we find that the qualitative performance of ALH is comparable to RAH. We do not model the overhead of creating the hierarchy nor the cost of adding router-assist to the network. Therefore, our conclusions inform rather than close the debate of which approach is better.
... This changes fundamentally how recovery mechanism can be designed as discussed in this paper. Recovering loss from local groups of users has also been discussed in [16], [17]. These groups are usually fixed or given without much control by the receivers. ...
We consider the delivery of reliable and streaming services using application-level multicast (ALM) by means of UDP, where packet loss has to be recovered via retransmission in a timely manner in order to offer high level of service. Since packets may be lost due to congestion, tree-reconfiguration or node failure, the traditional "vertical" recovery, whereby upstream nodes retransmit the lost packet is no longer effective. We therefore propose and investigate lateral error recovery (LER). In LER, hosts are divided into a number of planes, each of which forms an independent ALM tree. Since the correlation of error among the planes is likely to be low, a node can effectively recover its error "laterally" from nearby nodes in other planes. We employ the technique of global network positioning (GNP) to map the hosts into a coordinate space and identify a set of close neighbors for error recovery by constructing a Voronoi diagram for each plane. We present centralized and distributed algorithm on how to construct the Voronoi diagrams. Using Internet-like topologies, we show via simulations that our system achieves low overheads in terms of relative delay penalty and physical link stress. For reliable service, lateral recovery greatly reduces the average recovery time as compared with vertical recovery schemes. For streaming applications, LER achieves much lower residual loss rate under a certain deadline constraint.
... application-level overlays is mainly individual efforts, leading to few standard and reusable protocols. Meanwhile, many network services for transport, routing, and management [4,11,12,14,22], which have been well designed in the past decade, are left under-leveraged. (3) At application level, it is hard to achieve strong isolation between an overlay and the rest of the Internet. ...
... There exist a large number of welldesigned network protocols which for some reason have not been widely adopted. Examples include IP multicast, scalable reliable multicast [11,14], IP anycast [12], and active networking [4,22]. There also exist protocols that are still in the initial stage of incremental deployment such as IPv6. ...
We propose a novel alternative to application-level overlays called VIOLIN, or Virtual Internetworking on OverLay INfrastructure. Inspired by recent advances in virtual machines, VIOLINs are virtual and isolated networks created on top of an overlay infrastructure such as PlanetLab. Entities in a VIOLIN include virtual routers, switches, and end-hosts, all implemented in software and hosted by physical overlay hosts. The salient features of VIOLIN include: (1) Each VIOLIN is a `virtual world' with its own IP address space. And its activities and communications are strictly confined within the VIOLIN. (2) All VIOLIN entities can be created, deleted, or migrated on-demand. (3) It provides a new playground to deploy, leverage, and evaluate value-added network services which are not widely deployed in the real Internet. An application can simply connect to a VIOLIN and leverage the network services provided. (4) It releases application developers from network service implementation details, resulting in easier application implementation and maintenance. We have designed and implemented a prototype of VIOLIN in PlanetLab.
... In this section, we present RMI -a formal model of the Scalable Reliable Multicast (SRM) protocol [1]. RMI precisely specifies the behavior of the basic version of SRM -more sophisticated versions involve adaptive and local recovery schemes [1,9]. We proceed by describing RMI's architecture, briefly present RMI's packet loss recovery scheme, and derive some constraints on RMI's parameters. ...
In this paper, we present a formal model of the reliable multicast service that ensures eventual packet delivery with, possibly, some timeliness guarantees. This model dictates precisely what it means to be a member of the reliable multicast group and which packets are guaranteed delivery to which members of the group. Moreover, it is reasonable, implementable, and broad; that is, it captures the intended behavior of a large collection of reliable multicast protocols. We also present a formal model of the Scalable Reliable Multicast (SRM) protocol [1]. We show that our model of SRM is safe, in the sense that it is a faithful implementation of our model of the reliable multicast service; that is, it may only deliver appropriate packets to each member of the reliable multicast group. We also show that, under certain constraints, the implementation is live, in the sense that it guarantees the timely delivery of the appropriate packets to the appropriate members of the reliable multicast group.
... The primitives presented here are intended to allow overlay networks to accomplish much more than a single form of multicast communication. [15,24,25,26,29], yet there is no consensus that any single protocol is appropriate for deployment in the world's routers. One problem plaguing such attempts is that there is no single agreed upon semantics for reliable multicast. ...
Overlay networks represent a flexible and deployable approach for applications to obtain new network semantics, but they suffer from some efficiency concerns. To support overlay networks efficiently, two new primitives are proposed for implementation in the network layer. Packet Reflection allows end hosts to request short-circuit packet routing and duplication in nearby routers. Path Painting allows multiple end hosts to determine where their disparate paths to a rendezvous point meet, in order to facilitate overlay topology building. Both primitives are incrementally deployable.
... Another way is creating groups for loss regions instead of for individual losses. This is suitable for the case where stable loss neighborhoods exist [96], [97]. 4) Logical tree-based approaches: Hierarchy-based (or tree-based) approaches organize receivers into a logical tree. ...
