Community-based VM placement framework

Filiposka, Sonja; Mishev, Anastas; Juiz, Carlos

doi:10.1007/s11227-015-1546-1

Cited by 18 publications

(10 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The proposal presented in this paper stems from an already tested framework for cloud resource management that uses the notion of virtual and physical hierarchical communities . The original cloud‐based framework is based on a two‐level hierarchical approach supported by the datacentre network topology and the interconnectivity between VMs in the user requested cloud services.…”

Section: Fog Resource Management Strategiesmentioning

confidence: 99%

“…We sum up the virtual to physical community matching procedure done by first defining a set of variables in Table that is used in the two level multidimentional optimisation method proposed by the framework …”

Section: Fog Resource Management Strategiesmentioning

confidence: 99%

“…The objective is to define the best possible match as the smallest physical community that can accommodate the complete virtual community under all fixed constraints, so it is defined an aggregate function

{scriptG}_{F false(a_{i j} P M_{j} false)} false(C o m_{i} false)

, and its equivalent for virtual communities, where F ( x ) is the specific aggregation function

{P M}_{i}^{'} = G_{F (a_{i j} P M_{j})} (C o m_{i}) = \sum_{j = 1}^{d} a_{i j} P M_{j}

{V M}^{'} = G_{F (V M)} (C S) .

The high‐level optimisation consists of finding the community

{P M}_{i}^{'}

that optimises the higher level placement function

F_{H L} false(P M_{i}^{'}, {V M}^{'} false)

, and then the second optimisation level is defined for setting the placement of individual VM s from the virtual community into the PM s of the selected community by this lower level placement function F LL ( Com i , CS ). The cloud‐based framework highly efficiently improves the network and energy performance, while consolidating cloud servers (more details on the full implementation can be found in the work of Filiposka et al).…”

Section: Fog Resource Management Strategiesmentioning

confidence: 99%

See 2 more Smart Citations

Mobile‐aware dynamic resource management for edge computing

Filiposka

Mishev

Gilly

2019

Trans Emerging Tel Tech

Self Cite

View full text Add to dashboard Cite

Shifting cloud computing capabilities close to the edge enables provisioning of low latency location‐based internet services that are adapted to user behaviour. However, this can be achieved neither with a simple move of the physical hosts closer to edge networks, nor continuing to abide by the same principles as the ones implemented in traditional cloud computing approaches. In order to accomplish the promised high quality of service, changes must be made to the resource management techniques so that they are adapted to the requirements of fog computing. This paper introduces a novel location‐based handoff management and its corresponding implementation of dynamic resource management modules that introduce resource allocation and migration strategies adapted specifically to fog computing. The aim is the implementation of the follow‐me behaviour of edge resources considering a tightly coupled perspective with the user location. The proposal is based on the concept of mapping physical areas to logical resource communities where the node mobility triggers migrations to the corresponding community. The results analysis from the simulation scenarios shows that the effectiveness of our community‐based dynamic resource management proposal in following the geographical trajectory of mobile users with wearable devices is over 80% even in highly saturated environments. The comparison with traditional resource management techniques clearly presents the advantages of our proposal, while the parameter wise in‐depth analysis discusses dependencies on the number of nodes, speed, and available resources.

show abstract

Section: Fog Resource Management Strategiesmentioning

confidence: 99%

Section: Fog Resource Management Strategiesmentioning

confidence: 99%

{scriptG}_{F false(a_{i j} P M_{j} false)} false(C o m_{i} false)

, and its equivalent for virtual communities, where F ( x ) is the specific aggregation function

{P M}_{i}^{'} = G_{F (a_{i j} P M_{j})} (C o m_{i}) = \sum_{j = 1}^{d} a_{i j} P M_{j}

{V M}^{'} = G_{F (V M)} (C S) .

The high‐level optimisation consists of finding the community

{P M}_{i}^{'}

that optimises the higher level placement function

F_{H L} false(P M_{i}^{'}, {V M}^{'} false)

Section: Fog Resource Management Strategiesmentioning

confidence: 99%

See 1 more Smart Citation

Mobile‐aware dynamic resource management for edge computing

Filiposka

Mishev

Gilly

2019

Trans Emerging Tel Tech

Self Cite

View full text Add to dashboard Cite

show abstract

“…A set of works previously applied complex network algorithms to the field of distributed systems. The use of topological metrics from the field of complex networks was proposed for the management of virtual machine (VM) placement in data centers 36 or for load balancing in cloud federations 37 . Later studies also proposed using complex network theory in the field of fog computing for the optimization of application placement and migration, 38 data storage, 5 or fog architecture organization 39 .…”

Section: Related Workmentioning

confidence: 99%

Optimization policy for file replica placement in fog domains

Guerrero

Lera

Juiz

2019

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

Fog computing architectures distribute computational and storage resources along the continuum from the cloud to things. Therefore, the execution of services or the storage of files can be closer to the users. The main objectives of fog computing domains are to reduce the user latency and the network usage. Availability is also an issue in fog architectures because the topology of the network does not guarantee redundant links between devices. Consequently, the definition of placement polices is a key challenge. We propose a placement policy for data replication to increase data availability that contrasts with other storage policies that only consider a single replica of the files. The system is modeled with complex weighted networks and topological features, such as centrality indices. Graph partition algorithms are evaluated to select the fog devices that store data replicas. Our approach is compared with two other placement policies: one that stores only one replica and FogStore, which also stores file replicas but uses a greedy approach (the shortest path). We analyze 22 experiments with simulations. The results show that our approach obtains the shortest latency times, mainly for writing operations, a smaller network usage increase, and a similar file availability to FogStore.

show abstract

“…Hence, an optimal design for a Data Center needs to take different parameters into account [13], such as a good tradeoff between data availability and security, as well as cost and energy efficiency, along with easiness to operability and maintainability. Likewise, redundant paths between any pair of hosts is a must when dealing with Data Centers [14], and therefore, an orchestrator might be used so as to choose just one path or some of them, whilst leaving the rest aside, hence activating the switches involved and leaving out the others, depending on the load balancing policies put in place or the energy consumption scheme being in use in the whole system [15], resulting in less energy usage overall, accounting for both power and cooling. Needless to say that if all redundant paths are active, all necessary switches will do so accordingly [16].…”

Section: Introductionmentioning

confidence: 99%