DQMP: A Decentralized Protocol to Enforce Global Quotas in Cloud Environments

Behl, Johannes; Distler, Tobias; Kapitza, Rüdiger

doi:10.1007/978-3-642-33536-5_21

Cited by 6 publications

(6 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this section, we revisit two works that inspire our vision for enforcing invariants without requiring coordination in the critical path of operation execution: the escrow transactional method [26] and the demarcation protocol [7]. We discuss the use of these protocol to provide the invariants from Section 2 without using strong consistency, or replica coordination in the general case.…”

Section: Old Revisitedmentioning

confidence: 99%

Towards Fast Invariant Preservation in Geo-replicated Systems

Balegas

Duarte

Ferreira

et al. 2015

SIGOPS Oper. Syst. Rev.

View full text Add to dashboard Cite

Today's global services and applications are expected to be highly available, scale to an unprecedented number of clients, and offer reliable, low-latency operations. This can be achieved through geo-replication, particularly when data consistency is relaxed. There are, however, applications whose data must obey global invariants at all times. Strong consistency protocols easily address this issue, but require global coordination among replicas and inevitably degrade application throughput and latency. While coordination is an inherent requirement for maintaining global application invariants, there are instances where coordination on a per operation basis can be avoided. In particular, it has been shown that either moving coordination outside the critical path for executing operations, or having one coordination round for multiple operations, are both effective ways to maintain global invariants and avoid most of the penalties of coordination. However, current georeplication protocols still have not taken advantage of these observations. In this paper, we review the design space of current solutions for building geo-replicated applications and present our guiding vision towards a general technique for providing global application invariants under eventual consistency, as a much cheaper alternative to strong consistency.

show abstract

Section: Old Revisitedmentioning

confidence: 99%

Towards Fast Invariant Preservation in Geo-replicated Systems

Balegas

Duarte

Ferreira

et al. 2015

SIGOPS Oper. Syst. Rev.

View full text Add to dashboard Cite

show abstract

“…Behl et al [2] present a distributed quota enforcement protocol. In this solution, free quotas are equally distributed between nodes running applications consumed by a given customer.…”

Section: Related Workmentioning

confidence: 99%

“…Theoretically, they are able to consume geographically dispersed resources unlimitedly. This could be risky since the cost of vast amounts of resources consumed unintentionally could be tremendous and may even exceed the estimated profit of the services [2]. To manage the services and limit global resource consumption, the providers may need to deploy multiple coordinators, and monitor the state of resource consumption.…”

Section: Introductionmentioning

confidence: 99%

Resource State Monitoring of Service Transactions in Cloud Systems

Zhang

Lakew

et al. 2014

2014 IEEE International Conference on Services Computing

View full text Add to dashboard Cite

Abstract-In cloud systems, services constituting a transaction may spread over a large number of servers or clusters. Theoretically, these services could consume cloud resources unlimitedly. To avoid financial loss due to resource overuse, clouds have to monitor the state of resources consumed by the services -collect values of consumption, and evaluate whether the combined usage of resources has excessed a pre-defined upper bound or not. The distributed nature of the services introduces a challenge to the monitoring system on how to summarise distributed state information with low cost. We present our resource state monitoring solution to capture the challenge introduced by services hosted in clouds. Our solution tracks the resource consumed by each service constituting a transaction individually whilst ensures the whole transaction does not overuse the allocated resource. It improves availability by avoiding single points of failure, and achieves scalability by minimising message exchanges. We performed experimental analyses that indicate this work can provide an inexpensive resource monitoring solution for transactions in clouds.

show abstract

“…It allows cluster administrators to tune the behavior of the protocol in response to failures between a reliable, heavyweight transaction protocol and an eventually consistent, lightweight coin recovery procedure. DQMP [4] combines a similar lightweight recovery protocol with a replicated, checkpointed management service to diffuse available resource capacity and balance local quotas across nodes. Neither system permits partitioned operation, and their resource balancing objectives may not be suitable for DRL, which tries to replicate max-min rate fairness between distributed TCP flows.…”

Section: Related Workmentioning

confidence: 99%

Scalable coordination of a tightly-coupled service in the wide area

Webb

Vattikonda

Yocum

et al. 2013

Proceedings of the First ACM SIGOPS Conference on Timely Results in Operating Systems

View full text Add to dashboard Cite

Today's large-scale services generally exploit looselycoupled architectures that restrict functionality requiring tight cooperation (e.g., leader election, synchronization, and reconfiguration) to a small subset of nodes. In contrast, this work presents a way to scalably deploy tightlycoupled distributed systems that require significant coordination among a large number of nodes in the wide area. Our design relies on a new group membership abstraction, circuit breaker, to preserve efficient pairwise communication despite transient link failures.We characterize our abstraction in the context of a distributed rate limiting (DRL) service's deployment on a global testbed infrastructure. Unlike most distributed services, DRL can safely operate in separate partitions simultaneously, but it requires timely snapshots of global state within each. Our DRL deployment leverages the circuit breaker abstraction along with a robust gossipbased communication protocol to meet its demanding communication requirements. Through local and widearea experiments, we illustrate that DRL remains accurate and fair in the face of a variety of failure scenarios.

show abstract

DQMP: A Decentralized Protocol to Enforce Global Quotas in Cloud Environments

Cited by 6 publications

References 17 publications

Towards Fast Invariant Preservation in Geo-replicated Systems

Towards Fast Invariant Preservation in Geo-replicated Systems

Resource State Monitoring of Service Transactions in Cloud Systems

Scalable coordination of a tightly-coupled service in the wide area

Contact Info

Product

Resources

About