No description available.

Distributed event processing systems suffer from poor scalability and inefficient resource usage caused by load distributions typical in real-world applications. The results of these shortcomings are availability issues, poor system performance, and high operating costs. This thesis proposes three remedies to solve these limitations in content-based publish/subscribe, which is a practical realization of an event processing system. First, we present a load balancing algorithm that relocates subscribers to distribute load and avoid overloads. Second, we propose publisher relocation algorithms that reduces both the load imposed onto brokers and delivery delay experienced by subscribers. Third, we present ``green" resource allocation algorithms that allocate as few brokers as possible while maximizing their resource usage efficiency by reconfiguring the publishers, subscribers, and the broker topology. We implemented and evaluated all of our approaches on an open source content-based publish/subscribe system called PADRES and evaluated them on SciNet, PlanetLab, a cluster testbed, and in simulations to prove the effectiveness of our solutions. Our evaluation findings are summarized as follows. One, the proposed load balancing algorithm is effective in distributing and balancing load originating from a single server to all available servers in the network. Two, our publisher relocation algorithm reduces the average input load of the system by up to 68%, average broker message rate by up to 85%, and average delivery delay by up to 68%. Three, our resource allocation algorithm reduces the average broker message rate even further by up to 92% and the number of allocated brokers by up to 91%.

Distributed event processing systems suffer from poor scalability and inefficient resource usage caused by load distributions typical in real-world applications. The results of these shortcomings are availability issues, poor system performance, and high operating costs. This thesis proposes three remedies to solve these limitations in content-based publish/subscribe, which is a practical realization of an event processing system. First, we present a load balancing algorithm that relocates subscribers to distribute load and avoid overloads. Second, we propose publisher relocation algorithms that reduces both the load imposed onto brokers and delivery delay experienced by subscribers. Third, we present ``green" resource allocation algorithms that allocate as few brokers as possible while maximizing their resource usage efficiency by reconfiguring the publishers, subscribers, and the broker topology. We implemented and evaluated all of our approaches on an open source content-based publish/subscribe system called PADRES and evaluated them on SciNet, PlanetLab, a cluster testbed, and in simulations to prove the effectiveness of our solutions. Our evaluation findings are summarized as follows. One, the proposed load balancing algorithm is effective in distributing and balancing load originating from a single server to all available servers in the network. Two, our publisher relocation algorithm reduces the average input load of the system by up to 68%, average broker message rate by up to 85%, and average delivery delay by up to 68%. Three, our resource allocation algorithm reduces the average broker message rate even further by up to 92% and the number of allocated brokers by up to 91%.

Abstract Peer-to-Peer (P2P) networks have been applied in many applications for sharing resources such as storage space, media files or network bandwidth. Their main benefits include decentralization, self-organization, and scalability. Moreover, P2P technologies are evolving towards hybrid systems, where P2P networks are used in those parts of a larger system to leverage the decentralization most efficiently. The examples include cloud computing, where P2P networks are used in sharing computing resources, and Machine-to-Machine communication, where P2P networks are used for resource discovery. In super-peer overlays, the nodes are either regular nodes or super nodes that are located higher in the node hierarchy. This type of overlay explores the heterogeneity of peers in the overlay network to enable applications to run more efficiently. Leveraging the advantage of a super-peer overlay for service provision is an important issue. This thesis contributes to the research and development of super-peer-based coordination service provision from three aspects. Firstly, a super-peer-based coordinated service provision framework is proposed to coordinate different service groups and service peers in resource sharing, aiming to enable service groups to adapt to dynamic service demands. The proposed framework is evaluated using the following performance metrics: service response time, scalability, robustness, and communication traffic, in comparison to related work. Secondly, an efficient algorithm for rapidly constructing a robust super-peer overlay is proposed. The algorithm introduces a super-peer candidate based method for super-peer selection and a two-hop search method for finding client peers. Performance evaluation takes into account the convergence time of building a super-peer overlay, communication overhead, scalability, robustness. A comparison to related work is also conducted. Thirdly, the architecture of distributed resource discovery based on P2P overlay for Machine-to-Machine service provision is proposed. The architecture supports heterogeneous devices using different communication protocols in resource registration and discovery for achieving interoperability. As a part of the thesis, a functional real-world prototype is implemented and verified with a simple demonstration. Preliminary evaluation on the prototype indicates that caching can improve the response time of resource lookup dramatically. / Tiivistelmä Vertaisverkkoja on hyödynnetty resurssien kuten tallennustilan, mediasisältöjen ja tietoliikennekapasiteetin jakamisessa. Niiden etuja perinteisiin keskitettyihin järjestelmiin verrattuna ovat hajautettu arkkitehtuuri, itseorganisoituvuus ja skaalautuvuus. Vertaisverkkoja käytetään yhä useammin järjestelmän osien toteuttamisessa, joissa hajautettujen resurssien hyödyntämisellä saavutetaan suurimmat edut. Esimerkkeinä ovat pilvilaskenta, jossa vertaisverkkoa käytetään laskentaresurssien jakamiseen, sekä laitteidenvälinen kommunikaatio, jossa vertaisverkkoja käytetään resurssien löytämiseen. Hierarkkisissa vertaisverkoissa niihin kytkeytyneet laitteet jaotellaan laitteiden kapasiteetin mukaan tavallisiin noodeihin ja näiden yläpuolella hierarkiassa toimiviin ylinoodeihin. Ylinoodeihin perustuva vertaisverkon kuoriverkko hyödyntää yksittäisten verkon noodien eli laitteiden erilaisuutta, jotta verkko voisi toimia tehokkaammin. Tämän ominaisuuden hyödyntäminen on erityisen tärkeää palvelun tarjonnassa. Tässä työssä on tutkittu ylinoodeihin perustuvan vertaisverkon palvelun tarjontaa kolmesta näkökulmasta. Ensimmäiseksi, työssä ehdotetaan ylinoodien koordinoimaa palveluntarjonnan toimintamallia resurssien jakamisessa. Toimintamallissa palveluryhmät ja palvelunoodit adaptoituvat dynaamisesti palvelupyyntöjen tarpeisiin. Tämän ratkaisun suorituskykyä arvioidaan palvelun vasteajan, skaalautuvuuden, robustisuuden ja tietoliikennemäärän suhteen verrattuna aiempiin ratkaisuihin. Toiseksi, työssä esitellään tehokas algoritmi robustin ylinoodikuoriverkon nopeaan muodostamiseen. Algoritmi käyttää ylinoodiehdokasmenetelmää ja kahden hypyn hakumetodia asiakasnoodien etsimisessä. Suorituskyvyn arvioinnissa otetaan huomioon ylinoodikuoriverkon konvergoitumisaika, tietoliikenneviestinnän aiheuttama ylimääräinen kuormitus, sekä järjestelmän skaalautuvuus ja robustisuus. Esitetyn algoritmin tehokkuutta arvioidaan vertaamalla näitä suorituskykymittareita aiempiin ratkaisuihin. Kolmanneksi, työssä esitellään hajautettu resurssihakemiston arkkitehtuuri, joka perustuu laitteiden välisen kuoriverkon palveluntarjontaan. Arkkitehtuuri tukee erilaisten laitteiden ja niiden käyttämien protokollien resurssien rekisteröintiä ja löytämistä yhteensopivuuden saavuttamiseksi. Väitöskirjatyön osana on toteutettu toimiva prototyyppi, jonka toimivuus on todennettu demonstraation avulla. Prototyypillä tehdyt mittaukset antavat perustellun syyn olettaa, että esitetyn ratkaisun mukainen välimuistin käyttö voi merkittävästi lyhentää resurssien etsimisen vasteaikaa.