Kjorveziroski, Vojdan

Preferred name

Official Name

Kjorveziroski, Vojdan

Main Affiliation

ORCID

0000-0003-0419-4300

Now showing 1 - 10 of 22

Some of the metrics are blocked by your
consent settings
Designing a Paas Solution for Large Scale Deployment of Student Projects
(IEEE, 2018-11)
;
;
;
Some of the metrics are blocked by your
consent settings
Анализа и дизајн на повеќенаменска архитектура на безсерверски платформи за транспарентни пресметки во облак-раб екосистемот
(ФИНКИ, УКИМ, Скопје, 2025)
The continuous advancements both in terms of software and hardware have made computing devices an omnipresent aspect of daily life, both for personal and business use. Miniaturization of hardware components has allowed every individual and corporation to be responsible for their own computing capacity through the use of personal computers, smartphones, and servers. Nevertheless, this approach requires continuous procurement of new and expensive devices to keep up with the pace of increasing computing demands. With the advent of cloud computing, these problems should have been overcome, allowing computing capacity to be available on demand, without requiring the possession of expensive equipment, thus transforming computing power into just another utility, such as electricity or tap water. Considering that the cloud is an abstract concept which at its core represents a physical infrastructure managed by a third party, many cloud providers offer ready-made services to provide more user-friendly interfaces. This approach also has significant drawbacks, mainly in terms of limited to non-existent compatibility between services offered by different cloud providers, resulting in a severe vendor lock-in effect. Additionally, cloud resources, in most cases, are billed no matter whether they are actually used or remain idle, leading to negative financial ramifications. Serverless computing is a new computing paradigm, with its most popular implementations currently limited to the cloud. Serverless computing has a real potential of realizing the vision of treating computing capacity just as another utility, while also reducing the overall cost, currently associated with other cloud services. Contrary to its name, servers are still involved, but they are completely transparent, allowing seamless provisioning and usage of software, without needing to take care of any infrastructure components. However, the full potential of serverless computing is still not completely realized, since many serverless solutions are hindered by subpar performance, limited interoperability, and restricted set of execution locations (either in the cloud or at the edge close to the end-users, but not both), resulting in a severe vendor lock-in. This doctoral thesis proposes a novel approach to serverless computing through analysis and design of multipurpose architecture for serverless platforms aimed at transparent execution in the edge-cloud continuum. The end-goal is the creation of such an edge-cloud continuum, an environment comprised of heterogenous infrastructures placed in different locations (both in the cloud and at the edge), owned by independent third parties, yet unified through a common interface. This thesis introduces a completely new execution environment for serverless functions and describes advanced connectivity strategies, allowing trusted communication between infrastructures no matter their geographical location or network conditions. The proposed approaches overcome the identified open issues in the area of serverless computing and act as foundational pillars to the architecture through which the vision of transparent computations across a unified edge-cloud continuum can finally be realized.
Some of the metrics are blocked by your
consent settings
Framework for a Multipurpose Remotely Accessible Laboratory for Education
(FINKI, 2022)
Hands-on exercises are an important aspect of students’ education and they have a positive effect on their overall knowledge retention rates. With the Covid-19 pandemic on one hand, and the ever-increasing number of new students on the other, in-person computer laboratories are no longer a feasible option for the implementation of such practical lessons.The introduction of remote learning laboratories is one potential solution to both of these problems. However, existing remote laboratory implementations only focus on either a single scientific field, are not versatile in terms of supported infrastructure, orsupport only limited runtime options for the software to be used as part of the exercises. To overcome these issues, we have devised a remote laboratory framework consisting of 8 fundamental feature requirements which would allow flexible use in different fields of study, while at the same time enabling easy extension with existing and new simulation software. We discuss different implementation routes and report on the progress that we have already made in terms of setting up such a remote laboratory for the needs of our courses. We conclude that while most features can be implemented with existing open source software,challenges arise in terms of providing a self-service portal from which students can on-demand deploy resources.
Some of the metrics are blocked by your
consent settings
Deploying Production-Grade Kubernetes Cluster
(2019-05)
;
Deploying a production-grade Kubernetes cluster is a challenging feat, mainly because of all the different services that need to be integrated with each other. We examined Kubespray, which is an open-source project whose purpose is to automate the deployment of stable clusters, as well as ease the future administration and lifecycle management along the way. Additional components can be installed during the deployment process that enrich and improve the cluster capabilities. We used Kubespray and together with other open-sources projects we created production-grade Kubernetes cluster ready for real deployment scenarios.
Some of the metrics are blocked by your
consent settings
On-Demand Network Management with NMaaS: Network Management as a Service
(IEEE, 2022-04-25)
;
Vuletic, Pavle V.
;
Lopatowski, Lukasz
;
Loui, Frederic
The need for network management is universal, no matter the size of the network. Unfortunately, monitoring is often burdensome for infrastructure teams, because despite catering to their production services, additional supporting systems need to be maintained, continuously updated, and configured. As a result of the ever-growing complexity of new network management solutions, the time required for testing a new tool is increasing, thus disincentivizing exploration of new alternatives. Network Management as a Service (NMaaS) is a software suite which allows teams to centrally manage multiple remote infrastructures through effortless, fast deployment and configuration of network management applications as well as supporting tools. Using the GitOps approach, all application configuration is versioned and automatically synced to running instances, allowing easy migration from existing installations, as well as roll-back of recent configuration changes. In this demonstration we outline the NMaaS architecture, discuss possible use-cases, and showcase the application deployment process together with the steps required for extending the existing application catalog.
Some of the metrics are blocked by your
consent settings
WebAssembly Orchestration in the Context of Serverless Computing
(Springer, 2023-07)
;
Recent WebAssembly advancements including better programming language support and the introduction of both the WebAssembly System Interface, and the WebAssembly Component Model, have transformed it from primarily a client-side technology to a server-side one as well. The advantages associated with WebAssembly, such as cross platform portability, small software artifacts sizes, fast start up times, and per execution isolation make it a good fit for serverless scenarios. While there are existing initiatives for using WebAssembly in such serverless contexts, orchestration is still an open question. To overcome this issue, we present a way for extending Kubernetes, allowing it to orchestrate natively executed WebAssembly modules, in addition to containers. We describe an extension to an existing WebAssembly software shim for containerd and a new Kubernetes WebAssembly operator. Benchmarking results from the proposed solution obtained using 9 serverless functions packaged both as WebAssembly modules and OpenFaaS functions running in containers, show that WebAssembly has clear advantages for frequently executed serverless functions which require elasticity. WebAssembly functions enjoy two times faster deployment times and at least an order of magnitude smaller artifact sizes while still offering comparable execution performance. However, when it comes to sustained performance for long running serverless functions with processor intensive workloads, containers are the preferred choice, compensating for the increased cold start times with faster execution times.
Some of the metrics are blocked by your
consent settings
Relocation of container-based services in a MEC-NFV orchestrated environment
(Springer, 2024)
Bernad, Cristina
;
;
Roig, Pedro
;
Alcaraz, Salvador
;
Gilly, Katja
With the rapid growth of real-time next-generation mobile services, it has become necessary to work towards holistic orchestration of the benefits promised with edge computing based on bringing the computing infrastructure closer to the end user. While the concept of Multi-access Edge Computing (MEC) integrated with Network Function Virtualisation (NFV) is being standardised, there is still a lot of work to be done to orchestrate the relocation of edge applications integrated in 5G and beyond systems in a smooth and efficient manner. In this paper, we document the current status of the transparent relocation of edge services in an experimentally deployed MEC-NFV environment based on OSM. Working towards gathering monitoring training datasets necessary for the development of proactive MEC application orchestrators that will implement seamless follow-me behaviour for MEC services, we provide benchmark results for the service downtime of three potential MEC services hosted in lightweight containers. Our analysis of results shows that containers exhibit improved performance over that of virtual machines, but there are still some issues that require improvement in both the orchestration implementation as well at the relocation process for containers.
Some of the metrics are blocked by your
consent settings
Implementing Multi-Access Edge Computing with Kubernetes
(ICT Innovations, 2022)
;
Bernad Canto, Cristina
;
Gilly, Katja
;
The introduction of novel telecommunication standards such as 5G and beyond, which offer increased throughput and more efficient network communication, serve as enablers for new technologies. One of these technologies is Multi-Access Edge Computing (MEC), with the potential to revolutionize existing computing architectures and their feature set as we know them to-day. By collocating compute nodes with mobile networking equipment, customers are offered reduced latency and increased privacy, compared to alternative scenarios where the traffic is indiscriminately routed to the cloud. However, the distributed nature of the mobile landscape, coupled with the huge number of involved parties, requires careful consideration and standardization before any rollout. A number of reference documents have been published in recent years with the aim of standardizing the communication interfaces of MEC. We discuss these standardization efforts and provide a description of a MEC architecture centered around the Kubernetes container orchestrator, replacing the concept of virtualized MEC applications with containerized instances which can be deployed as serverless functions. We also offer a use-case scenario leveraging the described component mapping, showcasing the scalability and load-balancing features of the proposed architecture.
Some of the metrics are blocked by your
consent settings
Evaluating IPv6 Support in Kubernetes
(IEEE, 2021-11-23)
;
;
Kubernetes is one of the most popular container orchestrators today, but IPv4 address scarcity and the lucrative reseller market introduce problems for cluster administrators who would like to deploy publicly available applications, without resorting to NAT. We discuss the extent of IPv6 support in recent versions of Kubernetes and offer solutions for IPv6 dual-stack implementation. We also provide a reference dual-stack IPv6 Kubernetes network architecture that can be employed in existing or new clusters.
Some of the metrics are blocked by your
consent settings
Serverless Platforms Performance Evaluation at the Network Edge
(Springer International Publishing, 2022)
;
;
Emerging computer paradigms aim to fulfill the ever-present ideal of running as many applications on existing infrastructure, as efficiently as possible. One such novel concept is serverless computing which abstracts away infrastructure management, scaling and deployment from developers, allowing them to host function instances with granular responsibilities. However, faced with the meteoric growth in the number of IoT devices, the cloud is no longer suitable to meet this demand and a shift to edge infrastructures is needed, providing reduced latencies. While there are existing serverless platforms, both commercial and open-source that can be deployed at the edge, a comprehensive performance analysis is needed to determine their advantages and drawbacks, define open-issues, and identify areas for improvement. This paper analyses three different serverless edge platforms with the help of an existing serverless test suite, outlining their architecture, as well as execution performance in both sequential and parallel invocation scenarios. Special focus is paid to solutions that can be deployed in a standalone fashion, without complex clustering requirements. Results show that while the serial execution performance is comparable among the analyzed platforms, there are noticeable differences in cases of concurrent executions.

Kjorveziroski, Vojdan

Filters

Settings

Sort By

Results per page