In this paper we present DMFE (did my function execute?), which is a concept capable of learning and recognizing functional-level events, states, and loads from low-level execution-data. DMFE-functions are not necessarily software functions, as in "my_fun( )", but general functions in the etymological sense of the word, such as "someone pushed code to git", or "player activity is high". This allows DMFE to act as a general multi-purpose sensor which can be applied across a variety of software components-to be used for software monitoring, debugging, or testing-all without requiring the need for a deep understanding of the source code. Since the truth is always in the code, the main idea behind DMFE is to have the code itself "paint" execution-data on a "canvas" during run-time, and then let a deep neural network detect patterns which it associates with these functions and behaviors. We have successfully applied DMFE on internal production-code, and to illustrate how this is done we have also applied it on the two open-source projects: i) the distributed version-control system Git and ii) a text-based multi-user dungeon game Mud.
{"title":"DMFE","authors":"Victor Millnert, Magnus Templing, Patrik Åberg","doi":"10.1145/3468737.3494086","DOIUrl":"https://doi.org/10.1145/3468737.3494086","url":null,"abstract":"In this paper we present DMFE (did my function execute?), which is a concept capable of learning and recognizing functional-level events, states, and loads from low-level execution-data. DMFE-functions are not necessarily software functions, as in \"my_fun( )\", but general functions in the etymological sense of the word, such as \"someone pushed code to git\", or \"player activity is high\". This allows DMFE to act as a general multi-purpose sensor which can be applied across a variety of software components-to be used for software monitoring, debugging, or testing-all without requiring the need for a deep understanding of the source code. Since the truth is always in the code, the main idea behind DMFE is to have the code itself \"paint\" execution-data on a \"canvas\" during run-time, and then let a deep neural network detect patterns which it associates with these functions and behaviors. We have successfully applied DMFE on internal production-code, and to illustrate how this is done we have also applied it on the two open-source projects: i) the distributed version-control system Git and ii) a text-based multi-user dungeon game Mud.","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128052621","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
E. Maleki, Weibin Ma, Lena Mashayekhy, Humberto J. La Roche
The demand for content such as multimedia services with stringent latency requirements has proliferated significantly, posing heavy backhaul congestion in mobile networks. The integration of Multi-access Edge Computing (MEC) and 5G network is an emerging solution that alleviates the backhaul congestion to meet QoS requirements such as ultra-low latency, ultra-high reliability, and continuous connectivity to support various latency-critical applications for user equipment (UE). Content caching can markedly augment QoS for UEs by increasing the availability of popular content. However, uncertainties originating from user mobility cause the most challenging barrier in deciding content routes for UEs that lead to minimum latency. Considering the 5G-enabled MEC components, it is critical to select the optimal 5G components, representing content routes from Edge Application Servers (EASs) to UEs, that enhances QoS for the UEs with uncertain mobility patterns by reducing frequent handover (path reallocation). To this aim, we study the component selection for QoS-aware content delivery in 5G-enabled MEC. We first formulate an integer programming (IP) optimization model to obtain the optimal content routing decisions. As this problem is NP-hard, we tackle its intractability by designing an efficient online learning approach, called Q-CSCD, to achieve a bounded performance. Q-CSCD learns the optimal component selection for UEs and autonomously makes decisions to minimize latency for content delivery. We conduct extensive experiments based on a real-world dataset to validate the effectiveness of our proposed algorithm. The results reveal that Q-CSCD leads to low latency and handover ratio in a reasonable time with a reduced regret over time.
{"title":"QoS-aware 5G component selection for content delivery in multi-access edge computing","authors":"E. Maleki, Weibin Ma, Lena Mashayekhy, Humberto J. La Roche","doi":"10.1145/3468737.3494101","DOIUrl":"https://doi.org/10.1145/3468737.3494101","url":null,"abstract":"The demand for content such as multimedia services with stringent latency requirements has proliferated significantly, posing heavy backhaul congestion in mobile networks. The integration of Multi-access Edge Computing (MEC) and 5G network is an emerging solution that alleviates the backhaul congestion to meet QoS requirements such as ultra-low latency, ultra-high reliability, and continuous connectivity to support various latency-critical applications for user equipment (UE). Content caching can markedly augment QoS for UEs by increasing the availability of popular content. However, uncertainties originating from user mobility cause the most challenging barrier in deciding content routes for UEs that lead to minimum latency. Considering the 5G-enabled MEC components, it is critical to select the optimal 5G components, representing content routes from Edge Application Servers (EASs) to UEs, that enhances QoS for the UEs with uncertain mobility patterns by reducing frequent handover (path reallocation). To this aim, we study the component selection for QoS-aware content delivery in 5G-enabled MEC. We first formulate an integer programming (IP) optimization model to obtain the optimal content routing decisions. As this problem is NP-hard, we tackle its intractability by designing an efficient online learning approach, called Q-CSCD, to achieve a bounded performance. Q-CSCD learns the optimal component selection for UEs and autonomously makes decisions to minimize latency for content delivery. We conduct extensive experiments based on a real-world dataset to validate the effectiveness of our proposed algorithm. The results reveal that Q-CSCD leads to low latency and handover ratio in a reasonable time with a reduced regret over time.","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121056201","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The benefits of cloud computing have attracted many organizations to migrate their IT infrastructures into the cloud. In an infrastructure as a service (IaaS) model, the cloud service provider offers services to multiple consumers using shared physical hardware resources. However, by sharing a cloud environment with other consumers, organizations may also share security risks with their cotenants. Distributed denial of service (DDoS) attacks are considered one of the major security threats in cloud computing. Without a proper defense mechanism, an attack against one tenant can also affect the availability of cotenants. This work uses a game-theoretic approach to analyze the interactions between various entities when the cloud is under attack. The resulting Nash equilibrium shows that collateral damage to cotenants is unlikely if the cloud service provider is unbiased and chooses a rational strategy, but the Nash equilibrium can change when the cloud service provider does not treat cloud consumers equally. The cloud service provider's bias can influence its strategy selection and create a situation where untargeted users suffer unnecessary collateral damage from DDoS attacks.
{"title":"Game-theoretic modeling of DDoS attacks in cloud computing","authors":"Kaho Wan, Joel Coffman","doi":"10.1145/3468737.3494093","DOIUrl":"https://doi.org/10.1145/3468737.3494093","url":null,"abstract":"The benefits of cloud computing have attracted many organizations to migrate their IT infrastructures into the cloud. In an infrastructure as a service (IaaS) model, the cloud service provider offers services to multiple consumers using shared physical hardware resources. However, by sharing a cloud environment with other consumers, organizations may also share security risks with their cotenants. Distributed denial of service (DDoS) attacks are considered one of the major security threats in cloud computing. Without a proper defense mechanism, an attack against one tenant can also affect the availability of cotenants. This work uses a game-theoretic approach to analyze the interactions between various entities when the cloud is under attack. The resulting Nash equilibrium shows that collateral damage to cotenants is unlikely if the cloud service provider is unbiased and chooses a rational strategy, but the Nash equilibrium can change when the cloud service provider does not treat cloud consumers equally. The cloud service provider's bias can influence its strategy selection and create a situation where untargeted users suffer unnecessary collateral damage from DDoS attacks.","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121246289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Containers provide a lightweight and fine-grained isolation for computational resources such as CPUs, memory, storage, and networks, but their weak isolation raises security concerns. As a result, research and development efforts have focused on redesigning truly sandboxed containers with system call intercept and hardware virtualization techniques such as gVisor and Kata Containers. However, such fully integrated sandboxing could overwhelm the lightweight and scalable nature of the containers. In this work, we propose a partially fortified sandboxing mechanism that concentratedly fortifies the network isolation, focusing on the fact that containerized clouds and the applications running on them require different isolation levels in accordance with their unique characteristics. We describe how to efficiently implement the mechanism to fortify network isolation for containers with a para-passthrough hypervisor and report evaluation results with benchmarks and real applications. Our findings demonstrate that this fortified network isolation has good potential to tailor sandboxes for containerized PaaS/FaaS clouds.
{"title":"Concentrated isolation for container networks toward application-aware sandbox tailoring","authors":"Yuki Nakata, Katsuya Matsubara, Ryosuke Matsumoto","doi":"10.1145/3468737.3494092","DOIUrl":"https://doi.org/10.1145/3468737.3494092","url":null,"abstract":"Containers provide a lightweight and fine-grained isolation for computational resources such as CPUs, memory, storage, and networks, but their weak isolation raises security concerns. As a result, research and development efforts have focused on redesigning truly sandboxed containers with system call intercept and hardware virtualization techniques such as gVisor and Kata Containers. However, such fully integrated sandboxing could overwhelm the lightweight and scalable nature of the containers. In this work, we propose a partially fortified sandboxing mechanism that concentratedly fortifies the network isolation, focusing on the fact that containerized clouds and the applications running on them require different isolation levels in accordance with their unique characteristics. We describe how to efficiently implement the mechanism to fortify network isolation for containers with a para-passthrough hypervisor and report evaluation results with benchmarks and real applications. Our findings demonstrate that this fortified network isolation has good potential to tailor sandboxes for containerized PaaS/FaaS clouds.","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127714935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In traditional cloud computing, dedicated hardware is substituted by dynamically allocated, utility-oriented resources such as virtualized servers. While cloud services are following the pay-as-you-go pricing model, resources are billed based on instance allocation and not on the actual usage, leading the customers to be charged needlessly. In serverless computing, as exemplified by the Function-as-a-Service (FaaS) model where functions are the basic resources, functions are typically not allocated or charged until invoked or triggered. Functions are not applications, however, and to build compelling serverless applications they frequently need to be orchestrated with some kind of application logic. A major issue emerging by the use of orchestration is that it complicates further the already complex billing model used by FaaS providers, which in combination with the lack of granular billing and execution details offered by the providers makes the development and evaluation of serverless applications challenging. Towards shedding some light into this matter, in this work we extensively evaluate the state-of-the-art function orchestrator AWS Step Functions (ASF) with respect to its performance and cost. For this purpose we conduct a series of experiments using a serverless data processing pipeline application developed as both ASF Standard and Express workflows. Our results show that Step Functions using Express workflows are economical when running short-lived tasks with many state transitions. In contrast, Standard workflows are better suited for long-running tasks, offering in addition detailed debugging and logging information. However, even if the behavior of the orchestrated AWS Lambda functions influences both types of workflows, Step Functions realized as Express workflows get impacted the most by the phenomena affecting Lambda functions.
{"title":"Exploring the cost and performance benefits of AWS step functions using a data processing pipeline","authors":"Anil Mathew, V. Andrikopoulos, F. Blaauw","doi":"10.1145/3468737.3494084","DOIUrl":"https://doi.org/10.1145/3468737.3494084","url":null,"abstract":"In traditional cloud computing, dedicated hardware is substituted by dynamically allocated, utility-oriented resources such as virtualized servers. While cloud services are following the pay-as-you-go pricing model, resources are billed based on instance allocation and not on the actual usage, leading the customers to be charged needlessly. In serverless computing, as exemplified by the Function-as-a-Service (FaaS) model where functions are the basic resources, functions are typically not allocated or charged until invoked or triggered. Functions are not applications, however, and to build compelling serverless applications they frequently need to be orchestrated with some kind of application logic. A major issue emerging by the use of orchestration is that it complicates further the already complex billing model used by FaaS providers, which in combination with the lack of granular billing and execution details offered by the providers makes the development and evaluation of serverless applications challenging. Towards shedding some light into this matter, in this work we extensively evaluate the state-of-the-art function orchestrator AWS Step Functions (ASF) with respect to its performance and cost. For this purpose we conduct a series of experiments using a serverless data processing pipeline application developed as both ASF Standard and Express workflows. Our results show that Step Functions using Express workflows are economical when running short-lived tasks with many state transitions. In contrast, Standard workflows are better suited for long-running tasks, offering in addition detailed debugging and logging information. However, even if the behavior of the orchestrated AWS Lambda functions influences both types of workflows, Step Functions realized as Express workflows get impacted the most by the phenomena affecting Lambda functions.","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126854701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Eric J. Lyons, Hakan Saplakoglu, M. Zink, Komal Thareja, A. Mandal, Chengyi Qu, Songjie Wang, P. Calyam, G. Papadimitriou, Ryan Tanaka, E. Deelman
Many Internet of Things (IoT) applications require compute resources that cannot be provided by the devices themselves. At the same time, processing of the data generated by IoT devices often has to be performed in real- or near real-time. Examples of such scenarios are autonomous vehicles in the form of cars and drones where the processing of observational data (e.g., video feeds) needs to be performed expeditiously to allow for safe operation. To support the computational needs and timeliness requirements of such applications it is essential to include suitable edge resources to execute these applications. In this paper, we present our FlyNet architecture which has the goal to provide a new platform to support workflows that include applications executing at the network edge, at the computing core, and leverage deeply programmable networks. We discuss the challenges associated with provisioning such networking and compute infrastructure on demand, tailored to IoT application workflows. We describe a strategy to leverage the end-to-end integrated infrastructure that covers all points in the spectrum of response latency for application processing. We present our prototype implementation of the architecture and evaluate its performance for the case of drone video analytics workflows with varying computational requirements.
{"title":"FlyNet","authors":"Eric J. Lyons, Hakan Saplakoglu, M. Zink, Komal Thareja, A. Mandal, Chengyi Qu, Songjie Wang, P. Calyam, G. Papadimitriou, Ryan Tanaka, E. Deelman","doi":"10.1145/3468737.3494098","DOIUrl":"https://doi.org/10.1145/3468737.3494098","url":null,"abstract":"Many Internet of Things (IoT) applications require compute resources that cannot be provided by the devices themselves. At the same time, processing of the data generated by IoT devices often has to be performed in real- or near real-time. Examples of such scenarios are autonomous vehicles in the form of cars and drones where the processing of observational data (e.g., video feeds) needs to be performed expeditiously to allow for safe operation. To support the computational needs and timeliness requirements of such applications it is essential to include suitable edge resources to execute these applications. In this paper, we present our FlyNet architecture which has the goal to provide a new platform to support workflows that include applications executing at the network edge, at the computing core, and leverage deeply programmable networks. We discuss the challenges associated with provisioning such networking and compute infrastructure on demand, tailored to IoT application workflows. We describe a strategy to leverage the end-to-end integrated infrastructure that covers all points in the spectrum of response latency for application processing. We present our prototype implementation of the architecture and evaluate its performance for the case of drone video analytics workflows with varying computational requirements.","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125282905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Anshul Jindal, Julian Frielinghaus, Mohak Chadha, M. Gerndt
With the advent of serverless computing in different domains, there is a growing need for dynamic adaption to handle diverse and heterogeneous functions. However, serverless computing is currently limited to homogeneous Function-as-a-Service (FaaS) deployments or simply FaaS Deployment (FaaSD) consisting of deployments of serverless functions using a FaaS platform in a region with certain memory configurations. Extending serverless computing to support Heterogeneous FaaS Deployments (HeteroFaaSDs) consisting of multiple FaaSDs with variable configurations (FaaS platform, region, and memory) and dynamically load balancing the invocations of the functions across these FaaSDs within a HeteroFaaSD can provide an optimal way for handling such serverless functions. In this paper, we present a software system called Courier that is responsible for optimally distributing the invocations of the functions (called delivering of serverless functions) within the HeteroFaaSDs based on the execution time of the functions on the FaaSDs comprising the HeteroFaaSDs. To this end, we developed two approaches: Auto Weighted Round-Robin (AWRR) and PerFunction Auto Weighted Round-Robin (PFAWRR) that use functions execution times for delivering serverless functions within a HeteroFaaSD to reduce the overall execution time. We demonstrate and evaluate the functioning of our developed tool on three HeteroFaaSDs using three FaaS platforms: 1) on-premise Open-Whisk, 2) AWS Lambda, and 3) Google Cloud Functions (GCF). We show that Courier can improve the overall performance of the invocations of the functions within a HeteroFaaSD as compared to traditional load balancing algorithms.
{"title":"Courier: delivering serverless functions within heterogeneous FaaS deployments","authors":"Anshul Jindal, Julian Frielinghaus, Mohak Chadha, M. Gerndt","doi":"10.1145/3468737.3494097","DOIUrl":"https://doi.org/10.1145/3468737.3494097","url":null,"abstract":"With the advent of serverless computing in different domains, there is a growing need for dynamic adaption to handle diverse and heterogeneous functions. However, serverless computing is currently limited to homogeneous Function-as-a-Service (FaaS) deployments or simply FaaS Deployment (FaaSD) consisting of deployments of serverless functions using a FaaS platform in a region with certain memory configurations. Extending serverless computing to support Heterogeneous FaaS Deployments (HeteroFaaSDs) consisting of multiple FaaSDs with variable configurations (FaaS platform, region, and memory) and dynamically load balancing the invocations of the functions across these FaaSDs within a HeteroFaaSD can provide an optimal way for handling such serverless functions. In this paper, we present a software system called Courier that is responsible for optimally distributing the invocations of the functions (called delivering of serverless functions) within the HeteroFaaSDs based on the execution time of the functions on the FaaSDs comprising the HeteroFaaSDs. To this end, we developed two approaches: Auto Weighted Round-Robin (AWRR) and PerFunction Auto Weighted Round-Robin (PFAWRR) that use functions execution times for delivering serverless functions within a HeteroFaaSD to reduce the overall execution time. We demonstrate and evaluate the functioning of our developed tool on three HeteroFaaSDs using three FaaS platforms: 1) on-premise Open-Whisk, 2) AWS Lambda, and 3) Google Cloud Functions (GCF). We show that Courier can improve the overall performance of the invocations of the functions within a HeteroFaaSD as compared to traditional load balancing algorithms.","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"144 8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129605487","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Future networks are expected to support cross-domain, cost-aware and fine-grained services in an efficient and flexible manner. Service Function Chaining (SFC) has been introduced as a promising approach to deliver these services. In the literature, centralized resource orchestration is usually employed to process SFC requests and manage computing and network resources. However, centralized approaches inhibit the scalability and domain autonomy in multi-domain networks. They also neglect location and hardware dependencies of service chains. In this paper, we propose federated service chaining, a distributed framework which orchestrates and maintains the SFC placement while sharing a minimal amount of domain information and control. We first formulate a deployment cost minimization problem as an Integer Linear Programming (ILP) problem with fine-grained constraints for location and hardware dependencies, which is NP-hard. We then devise a Distributed Federated Service Chaining placement approach (DFSC) using inter-domain paths and border nodes information. Our extensive experiments demonstrate that DFSC efficiently optimizes the deployment cost, supports domain autonomy and enables faster decision-making. The results show that DFSC finds solutions within a factor 1.15 of the optimal solution. Compared to a centralized approach in the literature, DFSC reduces the deployment cost by 12% while being one order of magnitude faster.
{"title":"Distributed federated service chaining for heterogeneous network environments","authors":"Chen Chen, Lars Nagel, Lin Cui, Fung Po Tso","doi":"10.1145/3468737.3494091","DOIUrl":"https://doi.org/10.1145/3468737.3494091","url":null,"abstract":"Future networks are expected to support cross-domain, cost-aware and fine-grained services in an efficient and flexible manner. Service Function Chaining (SFC) has been introduced as a promising approach to deliver these services. In the literature, centralized resource orchestration is usually employed to process SFC requests and manage computing and network resources. However, centralized approaches inhibit the scalability and domain autonomy in multi-domain networks. They also neglect location and hardware dependencies of service chains. In this paper, we propose federated service chaining, a distributed framework which orchestrates and maintains the SFC placement while sharing a minimal amount of domain information and control. We first formulate a deployment cost minimization problem as an Integer Linear Programming (ILP) problem with fine-grained constraints for location and hardware dependencies, which is NP-hard. We then devise a Distributed Federated Service Chaining placement approach (DFSC) using inter-domain paths and border nodes information. Our extensive experiments demonstrate that DFSC efficiently optimizes the deployment cost, supports domain autonomy and enables faster decision-making. The results show that DFSC finds solutions within a factor 1.15 of the optimal solution. Compared to a centralized approach in the literature, DFSC reduces the deployment cost by 12% while being one order of magnitude faster.","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"151 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115104768","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Amoeba","authors":"Antonis Papaioannou, K. Magoutis","doi":"10.2307/j.ctv6wgf4q.56","DOIUrl":"https://doi.org/10.2307/j.ctv6wgf4q.56","url":null,"abstract":"","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"429 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123148919","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Luis Augusto Dias Knob, C. Kayser, Paulo S. Souza, T. Ferreto
Edge Computing emerged as a solution to new applications, like augmented reality, natural language processing, and data aggregation that relies on requirements that the Cloud does not entirely fulfill. Given that necessity, the application deployment in Edge scenarios usually uses container-based virtualization. When deployed in a resource-constrained infrastructure, the deployment latency to instantiate a container can increase due to bandwidth limitation or bottlenecks, which can significantly impact scenarios where the edge applications have a short life period, high mobility, or interdependence between different microservices. To attack this problem, we propose a novel container scheduler based on a multi-objective genetic algorithm. This scheduler has the main objective of ensuring the Service Level Agreement set on each application that defines when the application is expected to be effectively active in the infrastructure. We also validated our proposal using simulation and evaluate it against two scheduler algorithms, showing a decrease in the number of applications that do not fulfill the SLA and the average time over the SLA to not fulfilled applications.
{"title":"Enforcing deployment latency SLA in edge infrastructures through multi-objective genetic scheduler","authors":"Luis Augusto Dias Knob, C. Kayser, Paulo S. Souza, T. Ferreto","doi":"10.1145/3468737.3494100","DOIUrl":"https://doi.org/10.1145/3468737.3494100","url":null,"abstract":"Edge Computing emerged as a solution to new applications, like augmented reality, natural language processing, and data aggregation that relies on requirements that the Cloud does not entirely fulfill. Given that necessity, the application deployment in Edge scenarios usually uses container-based virtualization. When deployed in a resource-constrained infrastructure, the deployment latency to instantiate a container can increase due to bandwidth limitation or bottlenecks, which can significantly impact scenarios where the edge applications have a short life period, high mobility, or interdependence between different microservices. To attack this problem, we propose a novel container scheduler based on a multi-objective genetic algorithm. This scheduler has the main objective of ensuring the Service Level Agreement set on each application that defines when the application is expected to be effectively active in the infrastructure. We also validated our proposal using simulation and evaluate it against two scheduler algorithms, showing a decrease in the number of applications that do not fulfill the SLA and the average time over the SLA to not fulfilled applications.","PeriodicalId":254382,"journal":{"name":"Proceedings of the 14th IEEE/ACM International Conference on Utility and Cloud Computing","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129610887","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：