A control theoretical approach to non-intrusive geo-replication for cloud services
(2016) 55th IEEE Conference on Decision and Control 2016 p.1649-1656- Abstract
Complete data center failures may occur due to disastrous events such as earthquakes or fires. To attain robustness against such failures and reduce the probability of data loss, data must be replicated in another data center sufficiently geographically separated from the original data center. Implementing geo-replication is expensive as every data update operation in the original data center must be replicated in the backup. Running the application and the replication service in parallel is cost effective but creates a trade-off between potential replication consistency and data loss and reduced application performance due to network resource contention. We model this trade-off and provide a control-theoretical solution based on Model... (More)
Complete data center failures may occur due to disastrous events such as earthquakes or fires. To attain robustness against such failures and reduce the probability of data loss, data must be replicated in another data center sufficiently geographically separated from the original data center. Implementing geo-replication is expensive as every data update operation in the original data center must be replicated in the backup. Running the application and the replication service in parallel is cost effective but creates a trade-off between potential replication consistency and data loss and reduced application performance due to network resource contention. We model this trade-off and provide a control-theoretical solution based on Model Predictive Control to dynamically allocate network bandwidth to accommodate the objectives of both replication and application data streams. We evaluate our control solution through simulations emulating the individual services, their traffic flows, and the shared network resource. The MPC solution is able to maintain a consistent performance over periods of persistent overload, and is quickly able to indiscriminately recover once the system return to a stable state. Additionally, the MPC balances the two objectives of consistency and performance according to the proportions specified in the objective function.
(Less)
- author
- Dürango, Jonas LU ; Tärneberg, William LU ; Tomas, Luis ; Tordsson, Johan ; Kihl, Maria LU and Maggio, Martina LU
- organization
- publishing date
- 2016-12-27
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- 2016 IEEE 55th Conference on Decision and Control, CDC 2016
- article number
- 7798502
- pages
- 8 pages
- publisher
- IEEE - Institute of Electrical and Electronics Engineers Inc.
- conference name
- 55th IEEE Conference on Decision and Control 2016
- conference location
- Las Vegas, NV, United States
- conference dates
- 2016-09-12 - 2016-09-14
- external identifiers
-
- scopus:85010791243
- ISBN
- 9781509018376
- DOI
- 10.1109/CDC.2016.7798502
- language
- English
- LU publication?
- yes
- id
- 6207f33c-58ab-4eaf-82fd-4ef45c76110d
- date added to LUP
- 2016-09-08 16:54:29
- date last changed
- 2024-03-07 11:53:01
@inproceedings{6207f33c-58ab-4eaf-82fd-4ef45c76110d, abstract = {{<p>Complete data center failures may occur due to disastrous events such as earthquakes or fires. To attain robustness against such failures and reduce the probability of data loss, data must be replicated in another data center sufficiently geographically separated from the original data center. Implementing geo-replication is expensive as every data update operation in the original data center must be replicated in the backup. Running the application and the replication service in parallel is cost effective but creates a trade-off between potential replication consistency and data loss and reduced application performance due to network resource contention. We model this trade-off and provide a control-theoretical solution based on Model Predictive Control to dynamically allocate network bandwidth to accommodate the objectives of both replication and application data streams. We evaluate our control solution through simulations emulating the individual services, their traffic flows, and the shared network resource. The MPC solution is able to maintain a consistent performance over periods of persistent overload, and is quickly able to indiscriminately recover once the system return to a stable state. Additionally, the MPC balances the two objectives of consistency and performance according to the proportions specified in the objective function.</p>}}, author = {{Dürango, Jonas and Tärneberg, William and Tomas, Luis and Tordsson, Johan and Kihl, Maria and Maggio, Martina}}, booktitle = {{2016 IEEE 55th Conference on Decision and Control, CDC 2016}}, isbn = {{9781509018376}}, language = {{eng}}, month = {{12}}, pages = {{1649--1656}}, publisher = {{IEEE - Institute of Electrical and Electronics Engineers Inc.}}, title = {{A control theoretical approach to non-intrusive geo-replication for cloud services}}, url = {{https://lup.lub.lu.se/search/files/22381970/cdc.pdf}}, doi = {{10.1109/CDC.2016.7798502}}, year = {{2016}}, }