Marcelo Pasin

Byzantine Fault-Tolerant MapReduce: Faults are Not Just Crashes

MapReduce is often used to run critical jobs such as scientific data analysis. However, evidence in the literature shows that arbitrary faults do occur and can probably corrupt the results of MapReduce jobs. MapReduce runtimes like Hadoop tolerate crash faults, but not arbitrary or Byzantine faults. We present a MapReduce algorithm and prototype that tolerate these faults. An experimental evaluation shows that the execution of a job with our algorithms uses twice the resources of the original Hadoop, instead of the 3 or 4 times more that would be achieved with the direct application of common Byzantine fault-tolerance paradigms. We believe this cost is acceptable for critical applications that require that level of fault tolerance.

The TClouds architecture: Open and resilient cloud-of-clouds computing

In this paper we give a preliminary overview of the TClouds architecture, which is devoted to providing incrementally high levels of security and dependability to cloud infrastructures, in an open, modular and versatile way, based on the cloud-of-clouds paradigm. The TClouds architecture serves these objectives essentially by re-using and reconfiguring the same basic components, for example, intrusion-tolerant protocols (e.g., BFT). We exemplify concrete instantiations of the architecture and report on recently published proof-of-concept prototypes that validate the architectures effectiveness.

Athapascan Runtime: Efficiency for Irregular Problems

This paper introduces a multi-threaded, portable, parallel programming runtime system, called Athapascan-Ob. It is designed to support the efficient processing of large irregular problems. Parallel machines are considered as generic networks of heterogeneous symmetric multiprocessors, with remote thread creation, message exchange and remote memory accesses. Athapascan-Ob can dynamically map fine granularity computations on such machines and overlap communication delays with computation. These features should help to perform better load sharing and achieve better performance.

StreamHub: a massively parallel architecture for high-performance content-based publish/subscribe

By routing messages based on their content, publish/subscribe (pub/sub) systems remove the need to establish and maintain fixed communication channels. Pub/sub is a natural candidate for designing large-scale systems, composed of applications running in different domains and communicating via middleware solutions deployed on a public cloud. Such pub/sub systems must provide high throughput, filtering thousands of publications per second matched against hundreds of thousands of registered subscriptions with low and predictable delays, and must scale horizontally and vertically. As large-scale application composition may require complex publications and subscriptions representations, pub/sub system designs should not rely on the specific characteristics of a particular filtering scheme for implementing scalability. In this paper, we depart from the use of broker overlays, where each server must support the whole range of operations of a pub/sub service, as well as overlay management and routing functionality. We propose instead a novel and pragmatic tiered approach to obtain high-throughput and scalable pub/sub for clusters and cloud deployments. We separate the three operations involved in pub/sub and leverage their natural potential for parallelization. Our design, named StreamHub, is oblivious to the semantics of subscriptions and publications. It can support any type and number of filtering operations implemented by independent libraries. Experiments on a cluster with up to 384 cores indicate that StreamHub is able to register 150 K subscriptions per second and filter next to 2 K publications against 100 K stored subscriptions, resulting in nearly 400 K notifications sent per second. Comparisons against a broker overlay solution shows an improvement of two orders of magnitude in throughput when using the same number of cores.

Secure Content-Based Routing Using Intel Software Guard Extensions

Content-based routing (CBR) is a powerful model that supports scalable asynchronous communication among large sets of geographically distributed nodes. Yet, preserving privacy represents a major limitation for the wide adoption of CBR, notably when the routers are located in public clouds. Indeed, a CBR router must see the content of the messages sent by data producers, as well as the filters (or subscriptions) registered by data consumers. This represents a major deterrent for companies for which data is a key asset, as for instance in the case of financial markets or to conduct sensitive business-to-business transactions. While there exists some techniques for privacy-preserving computation, they are either prohibitively slow or too limited to be usable in real systems. In this paper, we follow a different strategy by taking advantage of trusted hardware extensions that have just been introduced in off-the-shelf processors and provide a trusted execution environment. We exploit Intels new software guard extensions (SGX) to implement a CBR engine in a secure enclave. Thanks to the hardware-based trusted execution environment (TEE), the compute-intensive CBR operations can operate on decrypted data shielded by the enclave and leverage efficient matching algorithms. Extensive experimental evaluation shows that SGX adds only limited overhead to insecure plaintext matching outside secure enclaves while providing much better performance and more powerful filtering capabilities than alternative software-only solutions. To the best of our knowledge, this work is the first to demonstrate the practical benefits of SGX for privacy-preserving CBR.

Elastic Scaling of a High-Throughput Content-Based Publish/Subscribe Engine

Publish/subscribe (pub/sub) infrastructures running as a service on cloud environments offer simplicity and flexibility for composing distributed applications. Provisioning them appropriately is however challenging. The amount of stored subscriptions and incoming publications varies over time, and the computational cost depends on the nature of the applications and in particular on the filtering operation they require (e.g., content-based vs. topic-based, encrypted vs. non-encrypted filtering). The ability to elastically adapt the amount of resources required to sustain given throughput and delay requirements is key to achieving cost-effectiveness for a pub/sub service running in a cloud environment. In this paper, we present the design and evaluation of an elastic content-based pub/sub system: E-STREAMHUB. Specific contributions of this paper include: (1) a mechanism for dynamic scaling, both out and in, of stateful and stateless pub/sub operators, (2) a local and global elasticity policy enforcer maintaining high system utilization and stable end-to-end latencies, and (3) an evaluation using real-world tick workload from the Frankfurt Stock Exchange and encrypted content-based filtering.

PAL: Exploiting Java Annotations for Parallelism

We discuss how Java annotations can be used to provide the meta information needed to automatically transform plain Java programs into suitable parallel code that can be run on workstation clusters, networks and grids. Programmers are only required to decorate the methods that will eventually be executed in parallel with standard Java 1.5 annotations. Then these annotations are automatically processed and parallel byte code is derived. When the annotated program is started, it automatically retrieves the information about the executing platform and evaluates the information specified inside the annotations to transform the byte-code into a semantically equivalent multithreaded or multitask version, depending on the target architecture features. The results returned by the annotated methods, when invoked, are futures with a wait-by-necessity semantics. A PAL (Parallel Abstraction Layer) prototype exploiting the annotation based parallelizing approach has been implemented in Java. PAL targets JJPF, an existing, skeleton based, JAVA/JINI programming environment, as Parallel Framework. The experiments made with the prototype are encouraging: the design of parallel applications has been greatly simplified and the performances obtained are the same of an application directly written in JJPF.

On the Feasibility of Byzantine Fault-Tolerant MapReduce in Clouds-of-Clouds

MapReduce is a framework for processing large data sets largely used in cloud computing. MapReduce implementations like Hadoop can tolerate crashes and file corruptions, but there is evidence that general arbitrary faults do occur and can affect the correctness of job executions. Furthermore, many individual cloud outages have been reported, raising concerns about depending on a single cloud. We present a MapReduce runtime that tolerates arbitrary faults and runs in a set of clouds at a reasonable cost in terms of computation and execution time. The main challenge is to avoid sending through the internet the huge amount of data that would normally be exchanged between map and reduce tasks.

On the Performance of Byzantine Fault-Tolerant MapReduce

MapReduce is often used for critical data processing, e.g., in the context of scientific or financial simulation. However, there is evidence in the literature that there are arbitrary (or Byzantine) faults that may corrupt the results of MapReduce without being detected. We present a Byzantine fault-tolerant MapReduce framework that can run in two modes: nonspeculative and speculative. We thoroughly evaluate experimentally the performance of these two versions of the framework, showing that they use around twice more resources than Hadoop MapReduce, instead of the three times more of alternative solutions. We believe this cost is acceptable for many critical applications.

Athapascan: An Experience on Mixing MPI Communications and Threads

Current parallel programming models as message passing exploit properly coarse-grain parallelism and suit well for regular applications. However, many applications have irregular behaviour and fine-grain parallelism, in which cases multithreading is more suitable. Multiprocessing and clustering have became cost-effective manner to build distributed-memory parallel machines due to technological progress. This paper discusses Athapascan, a multithreaded, portable, parallel programming runtime system, targeted for irregular applications. It is designed to integrate multithreading and communication, taking profit of both multiprocessing and communicating networks.