Altoros Releases Latest Comparative Performance Report of Three NoSQL Databases: Couchbase Server v6.6.0, MongoDB v4.2.11, and DataStax Enterprise v6.8.3 (Cassandra)

New study by Altoros analyzes the latency and throughput performance of each database using Yahoo! Cloud Serving Benchmark and custom tests

Altoros, a consultancy focusing on research and development for Global 2000 organizations, today announced the results of its latest performance benchmark report. The study provides a comparative analysis of the performance of three NoSQL databases: Couchbase Server v6.6.0, MongoDB v4.2.11 and DataStax Enterprise v6.8.3 (Cassandra). The relative performance of the databases were evaluated in terms of latency and the throughput they were capable of achieving. Moreover, the evaluation was conducted using three different cluster configurations - 4, 10, and 20 nodes - as well as four different workloads.

NoSQL encompasses a wide variety of database technologies that were developed in response to a rise in global volumes of data and the frequency with which this data is accessed. In contrast, relational databases were not designed to cope with the scalability and agility challenges that modern applications face, nor were they built to take advantage of the inexpensive storage and processing power available today. New-generation NoSQL systems help to achieve the highest levels of performance and uptime for workloads.

This report measures the relative performance in terms of latency and throughput that each database can achieve. The Yahoo! Cloud Serving Benchmark (YCSB) is an open-source specification and program suite for evaluating retrieval and maintenance capabilities of computer programs. It was used as the default tool for evaluation consistency. 

Workloads pressure

The first workload performs under an update-heavy mode - similar to a stock trading application - invoking 50% of reads and 50% of updates. The second workload performs a short-range scan that invokes 95% of scan and 5% of updates, where short ranges of records are queried instead of the individual ones. As such, the second workload simulates activities typical for an e-commerce application. The third workload represents a query with a single filtering option to which an offset and a limit are applied. Finally, the fourth workload is a join query with grouping and ordering applied.

Varying requirements

Altoros defined the database performance of the report by the speed at which the database processed basic operations. The basic operation is an action performed by a workload executor, which drives multiple client threads. Each thread executes a sequential series of operations by making calls to a database interface layer both to load a database (the load phase) and to execute a workload (the transaction phase).

The threads throttle the rate at which they generate requests so that Altoros can directly control the offered load against the database. Additionally, the threads measure latency and the achieved throughput of their operations and report these measurements to the statistics collection module.

"Based on our tests, Couchbase scales better than MongoDB on larger clusters. Couchbase uses a peer-to-peer structure, enabling direct access to nodes. Meanwhile, MongoDB has master-slave relationships, where certain operations have to call Mongoose, an Object Document Mapper, and a configuration server to access a node," said Artsiom Yudovin, Lead Data Engineer, Altoros.

Download the report here.

About Altoros

Altoros is an experienced IT services provider that helps enterprises to increase operational efficiency and accelerate the delivery of innovative products by shortening time to market. Relying on the power of cloud automation, microservices, AI/ML, and industry knowledge, its customers are able to get a sustainable competitive advantage. For more, visit www.altoros.com or follow @altoros.

# # #

Media contact

Alex Khizhniak

Altoros

+1 (650) 265-2266

pr@altoros.com

Source: Altoros

Share:


Categories: Business Technology

Tags: Altoros, Couchbase, DynamoDB, MongoDB, NoSQL