Empirical analysis of database server scalability using an N-tier benchmark with read-intensive workload

Abstract

The performance evaluation of database servers in N-tier applications is a serious challenge due to requirements such as non-stationary complex workloads and global consistency management when replicating database servers. We conducted an experimental evaluation of database server scalability and bottleneck identification in N-tier applications using the RUBBoS benchmark. Our experiments are comprised of a full scale-out mesh with up to nine database servers and three application servers. Additionally, the fourtier system was run in a variety of configurations, including two database management systems (MySQL and PostgreSQL), two hardware node types (normal and low-cost), and two database replication techniques (C-JDBC and MySQL Cluster). In this paper we present the analysis of results generated with a read-intensive interaction pattern (browse-only workload) in the client emulator. These empirical data can be divided into two kinds. First, for a relatively small number of servers, we find simple hardware resource bottlenecks. Consequently, system throughput increases with an increasing number of database (and application) servers. Second, when sufficient hardware resources are available, non-obvious database related bottlenecks have been found that limit system throughput. While the first kind of bottlenecks shows that there are similarities between database and application/web server scalability, the second kind of bottlenecks shows that database servers have significantly higher sophistication and complexity that require in-depth evaluation and analysis.

Performance Modelling and Simulation of Three-Tier Applications in Cloud and Multi-Cloud Environments

“…Moreover, previous research in multi-tier applications shows that inter-tier network performance typically does not cause significant performance penalty compared to CPU, RAM and I/O utilisation [51,52]. In fact Lloyd et al have empirically shown that for a multi-tier application in a cloud environment, the outgoing and incoming VM network transfer between tiers explain respectively only 0.75% and 0.74% of the overall performance variability.…”

Querying a graph database – language selection and performance considerations

“…We did not consider using an asynchronous event-driven implementation instead [58] that might positively impact performance when carefully tuned [59].…”

Automatic provisioning of multi-tier applications in cloud computing environments

“…Previous works [33] have demonstrated that inter-tier network performance typically does not cause significant overheads compared to processor, memory and I/O resources. That is the reason why the Informer module has not been initially designed to gather information about network latencies too.…”