Ingrid Jansch-Pôrto

Enforcing Fairness in a Live-Streaming System

We describe a practical auditing approach designed to encourage fairness in peer-to-peer streaming. Auditing is employed to ensure that correct nodes are able to receive streams even in the presence of nodes that do not upload enough data (opportunistic nodes), and scales well when compared to previous solutions that rely on tit-for-tat style of data exchange. Auditing involves two roles: local and global. Untrusted local auditors run on all nodes in the system, and are responsible for collecting and maintaining accountable information regarding data sent and received by each node. Meanwhile, one or more trusted global auditors periodically sample the state of participating nodes, estimate whether the streaming quality is satisfactory, and decide whether any actions are required. We demonstrate through simulation that our approach can successfully detect and react to the presence of opportunistic nodes in streaming sessions. Furthermore, it incurs low network and computational overheads, which remain fixed as the system scales.

A Network-Level Distributed Fault Injector for Experimental Validation of Dependable Distributed Systems

The use of Java for distributed systems and high-available applications demands the validation of their fault tolerance mechanisms to avoid unexpected behavior during execution. We present an extension of FIONA (fault injector oriented to network applications), a fault injection environment to experimentally validate the dependability of distributed Java applications. The main features of this extension are: its distributed architecture, which allows centralized configuration of multiple faults scenarios, and the support for a wider fault model associated with distributed systems, which includes network partitioning. For monitoring, FIONA supports the collection of log information and includes a helper application to integrate this information in a global log for post-mortem dependability analysis. FIONA is simple to operate, and we expect it to facilitate the conduction of validation experiments by application developers and testers

Implementing rollback-recovery coordinated checkpoints

Recovering from processor failures in distributed systems is an important problem in the design of reliable systems. The processes should coordinate their operation to guarantee that the set of local checkpoints taken by the individual processes form a consistent global checkpoint (recovery line). This allows the system to resume operation from a consistent global state, when recovering from failure. This paper shows the results of the implementation of a transparent (no special needs for applications) and coordinated (non blocking) rollback-recovery distributed algorithm. As it does not block applications, the overhead is reduced during failure-free operation. Furthermore, the rollback procedure can be executed fast as a recovery line is always available and well identified. Our preliminary experimental results show that the algorithm causes very low overhead on the performance (less than 2%), and high dependency on the checkpoint size. Now we study optimizations on the implementation to reduce checkpoint latency.

Challenging the feasibility of authentication mechanisms for P2P live streaming

Peer-to-peer live streaming applications are becoming increasingly popular, and already reach millions of users worldwide. There are, however, many aspects in which these applications must evolve, among which video quality and security deserve particular attention. While the former is about to become a reality given recent advances in codecs and link bandwidth, the latter still presents great challenges. The decentralized nature of P2P live streaming systems makes them vulnerable to attacks, among which pollution is arguably the most harmful. Multimedia stream authentication allows the detection of polluted content and can help identifying malicious users, but its costs may be prohibitive. This paper presents an analytical evaluation of recent authentication mechanisms in light of their application in P2P live streaming. Its key contribution is to compare, unlike previous work, overheads and security of various signature amortization and light digital signature mechanisms for P2P live streaming of standard and high definition video. We demonstrate, contrary to previous work, that current authentication mechanisms do not scale in therms of quality, and are feasible in current hardware only with a favorable set of parameters. Our results justify the investigation of more efficient and adaptable approaches to the next generation of P2P streaming with secure, high definition video.

Generating non-uniform distributions for fault injection to emulate real network behavior in test campaigns

Fault injection is an efficient technique to evaluate the robustness of computer systems and their fault tolerance strategies. In order to obtain accurate results from fault injection based tests, it is important to mimic real conditions during a test campaign. When testing dependability attributes of network applications the real faulty behavior of networks must be closely emulated. We show how probability distributions can be used to inject communication faults that closely resemble the behavior observed in real network environments. To demonstrate the strengths of this strategy we develop a reusable and extensible entity called FIEND, integrate it to a fault injector and use the resulting tool to run test experiments injecting non-uniform distributed faults in a network application taken as example.

Third workshop on theses and dissertations on dependable computing

The Workshop on Theses and Dissertations on Dependable Computing is a student forum for bringing together graduate students that research on topics related to dependable computing. The aim of this meeting is to present and discuss the proposed contribution, preliminary results and possible directions for their research. The previous editions of this Workshop were held in Florianopolis in conjunction with the Brazilian Symposium on Fault Tolerance (SCTF 2001), and in 2003 in Sao Paulo with the Latin-American Symposium on Dependable Computing (LADC 2003).

Improving Transputer Internal Memory Reliability with Transparent Test: Case Study and Performance Analysis

Abstract This paper presents a test procedure allowing on-line fault detection for the transputer internal memory during periodical field tests. The technique is based on Transparent BIST, which was implemented by software (transputer instruction set) and by taking advantage of the existing transputer hardware characteristics. The main characteristics of the transputer and internal memory are presented, as well as a detailed description of the test implementation and experimental results. Additionally, the degradation induced by the periodical field test in global system performance is discussed. The test approach can also be applied at the board level in order to verify the main memory system. Aiming to improve the performance of this type of test, it is also described some features being developed along with the PISH Project (Projeto Integrado de Software/Hardware - CNPq - Protem/CC). These features deal with the use of programmable logic devices (FPGAs) to implement dedicated hardware for board-level testing purposes.