Abstract
In the virtualization environment, a hypervisor scheduler determines the degree of shared resource occupancy of the virtual machine (VM) according to the degree of CPU processing and it provides a fair CPU processing on virtual machines (VMs). But VMs are experiencing unfair network performance due to the hypervisor scheduler’s policy occupying resources based on CPU processing time. In this paper, we present NIOECM which is a network IO event control technique that controls the network-intensive VM’s network IO event to guarantee a fairness network performance of VMs which have the same resource configuration. The NIOECM performs a network delay processing on the network-intensive VMs which have a high network I/O event set. As a result, the network-intensive VMs which have a low network I/O event set will have more chance to occupy the network resource. In the result of experiments, our approach provides more fairness of network performance and does not give a performance interference on VM which is performing another task.
| Original language | English |
|---|---|
| Title of host publication | Internet and Distributed Computing Systems 12th International Conference, IDCS 2019, Proceedings |
| Editors | Raffaele Montella, Angelo Ciaramella, Giancarlo Fortino, Antonio Guerrieri, Antonio Liotta |
| Publisher | Springer |
| Pages | 271-283 |
| Number of pages | 13 |
| ISBN (Print) | 9783030349134 |
| DOIs | |
| Publication status | Published - 2019 Jan 1 |
| Event | 12th International Conference on Internet and Distributed Computing Systems, IDCS 2019 - Naples, Italy Duration: 2019 Oct 10 → 2019 Oct 12 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 11874 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Conference
| Conference | 12th International Conference on Internet and Distributed Computing Systems, IDCS 2019 |
|---|---|
| Country | Italy |
| City | Naples |
| Period | 19/10/10 → 19/10/12 |
Fingerprint
Keywords
- Bandwidth
- Cloud
- Fairness
- Hypervisor
- Mitigation
- Network I/O
- Virtualization
- Workload aware
ASJC Scopus subject areas
- Theoretical Computer Science
- Computer Science(all)
Cite this
NIOECM : A network I/O event control mechanism to provide fairness of network performance among vms with same resource configuration in virtualized environment. / Lee, Jaehak; Kang, Jihun; Yu, Heonchang.
Internet and Distributed Computing Systems 12th International Conference, IDCS 2019, Proceedings. ed. / Raffaele Montella; Angelo Ciaramella; Giancarlo Fortino; Antonio Guerrieri; Antonio Liotta. Springer, 2019. p. 271-283 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11874 LNCS).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - NIOECM
T2 - A network I/O event control mechanism to provide fairness of network performance among vms with same resource configuration in virtualized environment
AU - Lee, Jaehak
AU - Kang, Jihun
AU - Yu, Heonchang
PY - 2019/1/1
Y1 - 2019/1/1
N2 - In the virtualization environment, a hypervisor scheduler determines the degree of shared resource occupancy of the virtual machine (VM) according to the degree of CPU processing and it provides a fair CPU processing on virtual machines (VMs). But VMs are experiencing unfair network performance due to the hypervisor scheduler’s policy occupying resources based on CPU processing time. In this paper, we present NIOECM which is a network IO event control technique that controls the network-intensive VM’s network IO event to guarantee a fairness network performance of VMs which have the same resource configuration. The NIOECM performs a network delay processing on the network-intensive VMs which have a high network I/O event set. As a result, the network-intensive VMs which have a low network I/O event set will have more chance to occupy the network resource. In the result of experiments, our approach provides more fairness of network performance and does not give a performance interference on VM which is performing another task.
AB - In the virtualization environment, a hypervisor scheduler determines the degree of shared resource occupancy of the virtual machine (VM) according to the degree of CPU processing and it provides a fair CPU processing on virtual machines (VMs). But VMs are experiencing unfair network performance due to the hypervisor scheduler’s policy occupying resources based on CPU processing time. In this paper, we present NIOECM which is a network IO event control technique that controls the network-intensive VM’s network IO event to guarantee a fairness network performance of VMs which have the same resource configuration. The NIOECM performs a network delay processing on the network-intensive VMs which have a high network I/O event set. As a result, the network-intensive VMs which have a low network I/O event set will have more chance to occupy the network resource. In the result of experiments, our approach provides more fairness of network performance and does not give a performance interference on VM which is performing another task.
KW - Bandwidth
KW - Cloud
KW - Fairness
KW - Hypervisor
KW - Mitigation
KW - Network I/O
KW - Virtualization
KW - Workload aware
UR - http://www.scopus.com/inward/record.url?scp=85075901185&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85075901185&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-34914-1_26
DO - 10.1007/978-3-030-34914-1_26
M3 - Conference contribution
AN - SCOPUS:85075901185
SN - 9783030349134
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 271
EP - 283
BT - Internet and Distributed Computing Systems 12th International Conference, IDCS 2019, Proceedings
A2 - Montella, Raffaele
A2 - Ciaramella, Angelo
A2 - Fortino, Giancarlo
A2 - Guerrieri, Antonio
A2 - Liotta, Antonio
PB - Springer
ER -