2016 IEEE International Conference on Smart Cloud (SmartCloud) 2016
DOI: 10.1109/smartcloud.2016.23
|View full text |Cite
|
Sign up to set email alerts
|

Intelligent Data Placement Mechanism for Replicas Distribution in Cloud Storage Systems

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
4
0
5

Year Published

2019
2019
2022
2022

Publication Types

Select...
6
2
2

Relationship

0
10

Authors

Journals

citations
Cited by 19 publications
(9 citation statements)
references
References 8 publications
0
4
0
5
Order By: Relevance
“…With that, it is possible to reduce the chances that the cluster becomes imbalanced. In general, examples of this approach, like the solution presented in [7], involve the development of new replica placement policies for the HDFS that consider the data volume stored in each node as criteria for data distribution. However, in some situations, like when new nodes are added to the system, it is not possible to prevent a high discrepancy between the data held on the cluster nodes, in a way that reactive approaches become necessary.…”
Section: Related Workmentioning
confidence: 99%
“…With that, it is possible to reduce the chances that the cluster becomes imbalanced. In general, examples of this approach, like the solution presented in [7], involve the development of new replica placement policies for the HDFS that consider the data volume stored in each node as criteria for data distribution. However, in some situations, like when new nodes are added to the system, it is not possible to prevent a high discrepancy between the data held on the cluster nodes, in a way that reactive approaches become necessary.…”
Section: Related Workmentioning
confidence: 99%
“…O balanceamento de réplicas no HDFS pode ser tratado de forma proativa ou reativa. Soluções proativas, tais como a de [Ibrahim et al 2016], atuam no momento da distribuição inicial dos blocos e envolvem a criação de novas políticas de posicionamento de réplicas para o HDFS que, preventivamente, de forma direta ou indireta, contribuem com uma distribuição de dados equilibrada. Embora estas soluções esforcem-se em manter o HDFS em um estado balanceado, nem sempreé possível impedir o desequilíbrio na distribuição de réplicas.…”
Section: Trabalhos Relacionadosunclassified
“…Um aspecto a ser considerado para o balanceamentoé o volume de dados armazenado em cada DN. Neste sentido, [Ibrahim et al 2016] propõem um novo algoritmo de posicionamento com objetivo de distribuir as réplicas de uma forma totalmente equilibrada entre os nodos. Para tal, baseado em suas cargas, os DNs são marcados como livres ou ocupados.…”
Section: Soluções Preventivasunclassified