The root server is at the top of the domain name hierarchical structure. To improve root service performance, each root deploys anycast nodes worldwide. What is the actual service performance of these nodes after deployment? We analyze the service performance of the root anycast nodes deployed in China based on the active measurement data detected by the VPs of different ISPs in different geographical locations. From the analysis, we find that the resolution performance of the roots with anycast nodes deployed in China is higher than that of roots without deployment. However, users of different operators have significant differences in accessing the root servers, such as parsing time, hitting anycast nodes, and most anycast nodes only providing services for one operator, limiting the service scope and reducing the service performance. The analysis results can help the root management and introduction institutions master the actual service status of the root servers, which can be used to optimize the performance of the existing root anycast nodes and provide a basis for deploying new root anycast nodes in the next step. Finally, we find that 67 top-level domain names are hijacked on the resolution path based on the measured data.

/pdf/performance-analysis-of-root-anycast-nodes-based-on-active-1gsvvcl5.pdf

Performance Analysis of Root Anycast Nodes Based on Active Measurement

Low-latency live streaming has imposed stringent latency requirements on video transport networks. In this paper we report on the design and operation of the Alibaba low-latency video transport network, LiveNet. LiveNet builds on a flat CDN overlay with a centralized controller for global optimization. As part of this, we present our design of the global routing computation and path assignment, as well as our fast data transmission architecture with fine-grained control of video frames. The performance results obtained from three years of operation demonstrate the effectiveness of LiveNet in improving CDN performance and QoE metrics. Compared with our prior state-of-the-art hierarchical CDN deployment, LiveNet halves the CDN delay and ensures 98% of views do not experience stalls and that 95% can start playback within 1 second. We further report our experiences of running LiveNet over the last 3 years.

LiveNet: a low-latency video transport network for large-scale live streaming

Today's networked and distributed applications, by and large, rely on cloud services. However, solely cloud-based services are not the ideal solution for all use cases, in particular, the case of high volume data sharing in scientific computing whose cloud usage costs could be prohibitively high. Thus we take on a task of building a distributed, federated data repository, dubbed Hydra, for sharing large volume scientific data. In this paper, we compare two design choices: designing Hydra over TCP/IP with a centralized controller, and designing Hydra over Named Data Network (NDN) to enable distributed control. Our study shows that (i) building Hydra over TCP/IP with a central controller offers a simple, straightforward design; (ii) however, the controller necessarily needs to be replicated for scalability and reliability, and cloud CDN is needed to scale data delivery, both bringing additional complexity into the overall design; and (iii) building Hydra over NDN automatically offers scalable and efficient data dissemination at volume, as well as enables distributed control with high resiliency.

Designing Hydra with Centralized versus Decentralized Control: A Comparative Study

Low-latency live streaming has imposed stringent latency requirements on video transport networks. In this paper we report on the design and operation of the Alibaba low-latency video transport network, LiveNet. LiveNet builds on a flat CDN overlay with a centralized controller for global optimization. As part of this, we present our design of the global routing computation and path assignment, as well as our fast data transmission architecture with fine-grained control of video frames. The performance results obtained from three years of operation demonstrate the effectiveness of LiveNet in improving CDN performance and QoE metrics. Compared with our prior state-of-the-art hierarchical CDN deployment, LiveNet halves the CDN delay and ensures 98% of views do not experience stalls and that 95% can start playback within 1 second. We further report our experiences of running LiveNet over the last 3 years. 

LiveNet

Knowledge of Internet paths allows operators and researchers to better understand the Internet and troubleshoot problems. Paths are often asymmetric, so measuring just the forward path only gives partial visibility. Despite the existence of Reverse Traceroute, a technique that captures reverse paths (the sequence of routers traversed by traffic from an arbitrary, uncontrolled destination to a given source), this technique did not fulfill the needs of operators and the research community, as it had limited coverage, low throughput, and inconsistent accuracy. In this paper we design, implement and evaluate revtr 2.0, an Internet-scale Reverse Traceroute system that combines novel measurement approaches and studies with a large-scale deployment to improve throughput, accuracy, and coverage, enabling the first exploration of reverse paths at Internet scale. revtr 2.0 can run 15M reverse traceroutes in one day. This scale allows us to open the system to external sources and users, and supports tasks such as traffic engineering and troubleshooting.

/pdf/internet-scale-reverse-traceroute-1qnjlye3.pdf

Internet scale reverse traceroute

The key to optimizing the performance of an anycast-based system (e.g., the root DNS or a CDN) is choosing the right set of sites to announce the anycast prefix. One challenge here is predicting catchments. A naive approach is to advertise the prefix from all subsets of available sites and choose the best-performing subset, but this does not scale well. We demonstrate that by conducting pairwise experiments between sites peering with tier-1 networks, we can predict the catchments that would result if we announce to any subset of the sites. We prove that our method is effective in a simplified model of BGP, consistent with common BGP routing policies, and evaluate it in a real-world testbed. We then present AnyOpt, a system that predicts anycast catchments. Using AnyOpt, a network operator can find a subset of anycast sites that minimizes client latency without using the naive approach. In an experiment using 15 sites, each peering with one of six transit providers, AnyOpt predicted site catchments of 15,300 clients with 94.7% accuracy and client RTTs with a mean error of 4.6%. AnyOpt identified a subset of 12 sites, announcing to which lowers the mean RTT to clients by 33ms compared to a greedy approach that enables the same number of sites with the lowest average unicast latency.

Zheyuan Zhang

Papers

AnyOpt: predicting and optimizing IP Anycast performance