RT-MongoDB: A NoSQL Database with Differentiated Performance

Andreoli, Remo; Cucinotta, Tommaso; Pedreschi, Dino

doi:10.5220/0010452400770086

Cited by 9 publications

(5 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This paper extends the modied version of MongoDB we presented in [2], improving the exibility in replicated scenarios where higher data durability is required. The key to ensure reliability and data consistency in MongoDB is to deploy a replica set, a group of multiple database instances residing on dierent physical machines that share the same data set, and then reply to a user write request only after a number of such instances have locally replicated the operation.…”

Section: Proposed Approachmentioning

confidence: 79%

“…This paper extends our prior work, RT-MongoDB [2], which exploits UNIX nice level combining them with two design choices of MongoDB, the per-client threading model and optimistic concurrency control, in order to achieve dierentiated performance. The additional modications we introduced pay special care to avoid penalizing the achievable throughput in presence of mixed-priority workloads, keeping eectiveness in separating performance among dierent-priority requests.…”

Section: Contributionsmentioning

confidence: 85%

“…The checkpoint system conceived in our previous work attempts to minimize this unbiased replication problem, by providing a prioritized channel to temporarily revoke the database access based on the user priority: low-priority, normal-priority or high-priority. In practice, the checkpoint system 2 As of MongoDB v4.4…”

Section: Proposed Approachmentioning

confidence: 99%

See 2 more Smart Citations

Differentiated Performance in NoSQL Database Access for Hybrid Cloud-HPC Workloads

Andreoli

Cucinotta

2021

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

In recent years, the demand for cloud-based high-performance computing applications and services has grown in order to sustain the computational and statistical challenges of big-data analytics scenarios. In this context, there is a growing need for reliable large-scale NoSQL data stores capable of eciently serving mixed high-performance and interactive cloud workloads. This paper deals with the problem of designing such NoSQL database service: to this purpose, a set of modications to the popular MongoDB software are presented. The modied MongoDB lets clients submit individual requests or even carry out whole sessions at dierent priority levels, so that the higher-priority requests are served with shorter response times that exhibit less variance, with respect to lower-priority requests. Experimental results carried out on two big multi-core servers using synthetic workload scenarios demonstrate the eectiveness of the proposed approach in providing dierentiated performance levels, highlighting what trade-os are available between maximum achievable throughput for the platform, and the response-time reduction for higher-priority requests.

show abstract

Section: Proposed Approachmentioning

confidence: 79%

Section: Contributionsmentioning

confidence: 85%

See 1 more Smart Citation

Differentiated Performance in NoSQL Database Access for Hybrid Cloud-HPC Workloads

Andreoli

Cucinotta

2021

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, the way WiredTiger is used within MongoDB makes busy waiting unsafe, as the rules recalled in Section I are not enforced. each remote connection to a MongoDB server is reserved a unique dedicated thread to handle the server-side activities [4], [5], which are ultimately performed by WiredTiger. Therefore, MongoDB should have controlled/restricted the number of clients concurrently issuing requests to avoid the potential waste of CPU time, as they result in threads concurrently using WiredTiger.…”

Section: A "Unsafe" Busy Waitingmentioning

confidence: 99%

Inducing Huge Tail Latency on a MongoDB deployment

Andreoli,

Cucinotta

2023

2023 IEEE International Conference on Cloud Engineering (IC2E)

Self Cite

View full text Add to dashboard Cite

The NoSQL paradigm has emerged as the leading design choice for cloud providers offering highly scalable storage services. Contrary to traditional relational databases, NoSQL architectures are capable of ingesting the ever-growing volume of nowadays' data-driven applications characterized by low-latency and high-throughput requirements. However, it is difficult to build an ultra-scalable, high-performance storage engine that can sustain an arbitrary number of concurrent clients. A common technique to increase throughput is minimizing the OS overhead, quantified as the number of context switches, through busy waiting (or "spinning"). While this simple synchronization mechanism proves to be beneficial in the high-performance computing community, it requires special care to avoid wasting resources and counter-intuitive behaviors. In this paper, we address an instance of "unsafe" busy waiting in WiredTiger, the underlying storage engine of MongoDB, which leads to a consistent, excessive increase of tail latency in high contention scenarios.

show abstract

“…At the service layer, there has been a similar focus on integrating time-criticality within the services' software. For instance, I worked on rt-MongoDB [3]- [5], a modified version of MongoDB which incorporates differentiated performance capabilities. If such feature get integrated in mainline, a cloud provider offering a cloud-hosted database service can use it to serve higher-priority requests with shorter response times and less variance, with respect to lower-priority requests.…”

Section: A Performance Predictabilitymentioning

confidence: 99%

Towards a Holistic Cloud System with End-to-End Performance Guarantees

Andreoli,

Cucinotta

2023

2023 IEEE International Conference on Cloud Engineering (IC2E)

Self Cite

View full text Add to dashboard Cite

Computing technologies are undergoing a relentless evolution from both the hardware and software sides, incorporating new mechanisms for low-latency networking, virtualization, operating systems, hardware acceleration, smart services orchestration, serverless computing, hybrid private-public Cloud solutions and others. Therefore, Cloud infrastructures are becoming increasingly attractive for deploying a wider and wider range of applications, including those with more and more stringent timing constraints, like the emerging use case of deploying time-critical applications. However, despite the availability of a number of public Cloud offerings, and of products (or open-source suites) for deploying in-house private Cloud infrastructures, still there are no solutions readily available for managing time-critical software components with predictable end-to-end timing requirements in the range of hundreds or even tens of milliseconds. The goal of this discussion is to present the multi-domain challenges associated with orchestrating a holistic Cloud system with endto-end guarantees, which is the subject of my current PhD investigations.

show abstract

RT-MongoDB: A NoSQL Database with Differentiated Performance

Cited by 9 publications

References 24 publications

Differentiated Performance in NoSQL Database Access for Hybrid Cloud-HPC Workloads

Differentiated Performance in NoSQL Database Access for Hybrid Cloud-HPC Workloads

Inducing Huge Tail Latency on a MongoDB deployment

Towards a Holistic Cloud System with End-to-End Performance Guarantees

Contact Info

Product

Resources

About