Gargamel: boosting DBMS performance by parallelising write transactions

Pierpaolo Cincilla 1, 2, * Sébastien Monnet 1, 2 Marc Shapiro 1, 2
* Corresponding author
1 LIP6 - Laboratoire d'Informatique de Paris 6
2 Regal - Large-Scale Distributed Systems and Applications
LIP6 - Laboratoire d'Informatique de Paris 6, Inria Paris-Rocquencourt
Abstract : Parallel transactions in distributed DBs incur high overhead for concurrency control and aborts. We propose an alternative approach by pre-serializing possibly conflicting trans- actions, and parallelizing non-conflicting update transactions to different replicas. Our system provides strong transactional guarantees. In effect, Gargamel partitions the database dynamically according to the update workload. Each database replica runs sequentially, at full bandwidth; mutual synchronisation between replicas remains minimal. Our simulations show that Gargamel improves both response time and load by an order of magnitude when contention is high (highly loaded system with bounded resources), and that otherwise slow-down is negligible
Keywords : Cloud computing distributed DBMS scheduling
Document type :
Conference papers
Complete list of metadatas

Cited literature [15 references]  Display  Hide  Download
https://hal.inria.fr/hal-00932795
Contributor : Marc Shapiro <>
Submitted on : Friday, January 17, 2014 - 4:38:53 PM
Last modification on : Thursday, March 21, 2019 - 2:41:40 PM
Long-term archiving on: Friday, April 18, 2014 - 11:50:45 AM

File

gargamel-ICPADS-2012.pdf
Files produced by the author(s)

Identifiers

Collections

CNRS | INRIA | UPMC | LIP6 | SORBONNE-UNIVERSITE | SU-SCIENCES

Citation

Pierpaolo Cincilla, Sébastien Monnet, Marc Shapiro. Gargamel: boosting DBMS performance by parallelising write transactions. ICPADS 2012 - IEEE 18th International Conference on Parallel and Distributed Systems, Dec 2012, Singapour, Singapore. pp.572--579, ⟨10.1109/ICPADS.2012.83⟩. ⟨hal-00932795⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

408

Files downloads

244

 

 

Contact

archive-ouverte@inria.fr