A fundamental problem that confronts peer-to-peer applications is to efficiently locate the node that stores a particular data item. This paper presents Chord, a distributed lookup protocol that addresses this problem. Chord provides support for just one operation: given a key, it maps the key onto a node. Data location can be easily implemented on top of Chord by associating a key with each data item, and storing the key/data item pair at the node to which the key maps. Chord adapts efficiently as nodes join and leave the system, and can answer queries even if the system is continuously changing. Results from theoretical analysis, simulations, and experiments show that Chord is scalable, with communication cost and the state maintained by each node scaling logarithmically with the number of Chord nodes.

Chord: A scalable peer-to-peer lookup service for internet applications

This paper presents the design and evaluation of Pastry, a scalable, distributed object location and routing substrate for wide-area peer-to-peer ap- plications. Pastry performs application-level routing and object location in a po- tentially very large overlay network of nodes connected via the Internet. It can be used to support a variety of peer-to-peer applications, including global data storage, data sharing, group communication and naming. Each node in the Pastry network has a unique identifier (nodeId). When presented with a message and a key, a Pastry node efficiently routes the message to the node with a nodeId that is numerically closest to the key, among all currently live Pastry nodes. Each Pastry node keeps track of its immediate neighbors in the nodeId space, and notifies applications of new node arrivals, node failures and recoveries. Pastry takes into account network locality; it seeks to minimize the distance messages travel, according to a to scalar proximity metric like the number of IP routing hops. Pastry is completely decentralized, scalable, and self-organizing; it automatically adapts to the arrival, departure and failure of nodes. Experimental results obtained with a prototype implementation on an emulated network of up to 100,000 nodes confirm Pastry's scalability and efficiency, its ability to self-organize and adapt to node failures, and its good network locality properties.

/pdf/pastry-scalable-decentralized-object-location-and-routing-1u2equbbbp.pdf

Pastry: Scalable, Decentralized Object Location, and Routing for Large-Scale Peer-to-Peer Systems

Hash tables - which map "keys" onto "values" - are an essential building block in modern software systems. We believe a similar functionality would be equally valuable to large distributed systems. In this paper, we introduce the concept of a Content-Addressable Network (CAN) as a distributed infrastructure that provides hash table-like functionality on Internet-like scales. The CAN is scalable, fault-tolerant and completely self-organizing, and we demonstrate its scalability, robustness and low-latency properties through simulation.

https://www.cs.sjtu.edu.cn/~gchen/course/acn/Reading/CAN.pdf

A scalable content-addressable network

A review of the geologic history of the Himalayan-Tibetan orogen suggests that at least 1400 km of north-south shortening has been absorbed by the orogen since the onset of the Indo-Asian collision at about 70 Ma. Significant crustal shortening, which leads to eventual construction of the Cenozoic Tibetan plateau, began more or less synchronously in the Eocene (50–40 Ma) in the Tethyan Himalaya in the south, and in the Kunlun Shan and the Qilian Shan some 1000–1400 km in the north. The Paleozoic and Mesozoic tectonic histories in the Himalayan-Tibetan orogen exerted a strong control over the Cenozoic strain history and strain distribution. The presence of widespread Triassic flysch complex in the Songpan-Ganzi-Hoh Xil and the Qiangtang terranes can be spatially correlated with Cenozoic volcanism and thrusting in central Tibet. The marked difference in seismic properties of the crust and the upper mantle between southern and central Tibet is a manifestation of both Mesozoic and Cenozoic tectonics. The form...

Geologic Evolution of the Himalayan-Tibetan Orogen

Reliability at massive scale is one of the biggest challenges we face at Amazon.com, one of the largest e-commerce operations in the world; even the slightest outage has significant financial consequences and impacts customer trust. The Amazon.com platform, which provides services for many web sites worldwide, is implemented on top of an infrastructure of tens of thousands of servers and network components located in many datacenters around the world. At this scale, small and large components fail continuously and the way persistent state is managed in the face of these failures drives the reliability and scalability of the software systems.This paper presents the design and implementation of Dynamo, a highly available key-value storage system that some of Amazon's core services use to provide an "always-on" experience. To achieve this level of availability, Dynamo sacrifices consistency under certain failure scenarios. It makes extensive use of object versioning and application-assisted conflict resolution in a manner that provides a novel interface for developers to use.

/pdf/dynamo-amazon-s-highly-available-key-value-store-2ozfk43ou4.pdf

Dynamo: amazon's highly available key-value store

In this paper, we propose the dissemination tree, a dynamic content distribution system built on top of a peer-to-peer location service. We present a replica placement protocol that builds the tree while meeting QoS and server capacity constraints. The number of replicas as well as the delay and bandwidth consumption for update propagation are significantly reduced. Simulation results show that the dissemination tree has close to the optimal number of replicas, good load distribution, small delay and bandwidth penalties for update multicast compared with the ideal case: static replica placement on IP multicast.

Dynamic Replica Placement for Scalable Content Delivery

The north-striking Tancheng-Lujiang (Tan-Lu) fault is a conspicuous and controversial feature of the eastern Asian landscape Near the southeast extremity of the fault in Anhui Province, we collected paleomagnetic samples at 17 Middle Triassic (T2) and 10 Upper Cretaceous (K2) to lower Cenozoic (E1) sites T2 remanent magnetizations are interpreted as primary in two of three areas The three areas are rotated 37° to 137° counterclockwise with respect to the South China Block (SCB) reference direction K2-E1 remanent magnetization directions pass regional fold and reversals tests and are not rotated with respect to surrounding areas Counterclockwise rotation of T2 strata therefore ended before K2 and is attributed to left lateral shear acting along Tan-Lu during the North China Block (NCB)-SCB collision In Shandong Province, 700 km north of the Anhui sites, four areas containing 33 Upper Jurassic (J3) and Cretaceous sites have negligible declination differences, except for one which has dispersed directions The fold test is inconclusive for this latter area and positive for the other three Regional concordance of the J3-E1 paleomagnetic data (including paleolatitudes) together with observed deformation patterns suggest that an extensional regime prevailed in the Late Cretaceous and Cenozoic Euler pole positions that constrain the North-South China collision and account for Tan-Lu motion suggest at least 500 km of sinistral shear took place along the fault, and either (1) subduction and related ultrahigh pressure (UHP) metamorphism occurred near the present location of the Qinling-Dabieshan and Sulu UHP belts while Tan-Lu acted as a transform fault that connected the two subduction zones, or (2) Tan-Lu and Sulu were parts of the same transform fault system and no UHP rocks formed in situ at Sulu In either case, UHP rocks originally exhumed near Dabieshan could have been transported by plate capture toward Sulu along Tan-Lu After North and South China impacted near Dabieshan, the Tan-Lu fault grew within the SCB as the Dabieshan corner indented the SCB, causing folds in SCB cover rocks to conform to the NCB margin Late Cretaceous to Cenozoic reactivation of Tan-Lu, with both right lateral strike-slip and normal fault motion, occurred as the SCB extruded east relative to the NCB under the influence of the India-Asia collision

https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/1999JB900123

Tectonic evolution of the Tancheng-Lujiang (Tan-Lu) fault via Middle Triassic to Early Cenozoic paleomagnetic data

Online social networks (OSNs) are exceptionally useful collaboration and communication tools for millions of users and their friends. Unfortunately, in the wrong hands, they are also extremely effective tools for executing spam campaigns and spreading malware.In this poster, we present an initial study to detect and quantitatively analyze the coordinated spam campaigns on online social networks in the wild. Our system detected about 200K malicious wall posts with embedded URLs, traced back to roughly 57K accounts. We find that more than 70% of all malicious wall posts are advertising phishing sites

/pdf/detecting-and-characterizing-social-spam-campaigns-4k0ftl7svs.pdf

Detecting and characterizing social spam campaigns

Overlay network monitoring enables distributed Internet applications to detect and recover from path outages and periods of degraded performance within seconds. For an overlay network with n end hosts, existing systems either require O(n2) measurements, and thus lack scalability, or can only estimate the latency but not congestion or failures. Our earlier extended abstract [1] briefly proposes an algebraic approach that selectively monitors k linearly independent paths that can fully describe all the O(n2) paths. The loss rates and latency of these k paths can be used to estimate the loss rates and latency of all other paths. Our scheme only assumes knowledge of the underlying IP topology, with links dynamically varying between lossy and normal.In this paper, we improve, implement and extensively evaluate such a monitoring system. We further make the following contributions: i) scalability analysis indicating that for reasonably large n (e.g., 100), the growth of k is bounded as O(n log n), ii) efficient adaptation algorithms for topology changes, such as the addition or removal of end hosts and routing changes, iii) measurement load balancing schemes, and iv) topology measurement error handling. Both simulation and Internet experiments demonstrate we obtain highly accurate path loss rate estimation while adapting to topology changes within seconds and handling topology errors.

/pdf/an-algebraic-approach-to-practical-and-scalable-overlay-542usoi58t.pdf

An algebraic approach to practical and scalable overlay network monitoring

Online social networks (OSNs) are extremely popular among Internet users Unfortunately, in the wrong hands, they are also effective tools for executing spam campaigns In this paper, we present an online spam filtering system that can be deployed as a component of the OSN platform to inspect messages generated by users in real-time We propose to reconstruct spam messages into campaigns for classification rather than examine them individually Although campaign identification has been used for offline spam analysis, we apply this technique to aid the online spam detection problem with sufficiently low overhead Accordingly, our system adopts a set of novel features that effectively distinguish spam campaigns It drops messages classified as “spam” before they reach the intended recipients, thus protecting them from various kinds of fraud We evaluate the system using 187 million wall posts collected from Facebook and 17 million tweets collected from Twitter In different parameter settings, the true positive rate reaches 809% while the false positive rate reaches 019% in the best case In addition, it stays accurate for more than 9 months after the initial training phase Once deployed, it can constantly secure the OSNs without the need for frequent re-training Finally, tested on a server machine with eight cores (Xeon E5520 22Ghz) and 16GB memory, the system achieves an average throughput of 1580 messages/sec and an average processing latency of 215ms on the Facebook dataset

/pdf/towards-online-spam-filtering-in-social-networks-20v3rangmd.pdf

Yan Chen

Papers

Dynamic Replica Placement for Scalable Content Delivery

Tectonic evolution of the Tancheng-Lujiang (Tan-Lu) fault via Middle Triassic to Early Cenozoic paleomagnetic data

Detecting and characterizing social spam campaigns

An algebraic approach to practical and scalable overlay network monitoring

Towards Online Spam Filtering in Social Networks