Blog, Events & Press

9 Jul

Scalability in Performance Management

Scalability is one of the most important criteria when evaluating Performance Management solutions. It represents more than the ability to monitor a large number of devices, objects, and metrics. Scalability is the capacity to do so while maintaining speed of information and keeping costs down.

A classical centralized database architecture hinders scalability of a performance monitoring platform. As shown in the diagram below, all the collectors feed performance data into one or more centralized databases. When an end user queries data for reports, the centralized database and report front end become a bottleneck. The more data collectors added to the system, the more congested the throughput. It’s like funneling the traffic from four major interstate highways through a single tollbooth. Everything comes to a halt. This is why reporting on the performance of large networks and data centers takes minutes or even hours.

Another expensive side effect of this architecture is the need for more human administration. Numerous servers, databases, operating systems, security updates and licensing add to the complexity. In the end, performance monitoring solutions built around a centralized database architecture become problematic. As monitoring needs expand, the solutions become expensive and fail at providing instantaneous information about the health of the environment.

Centralized Database Scalability In SevOne

A centralized database architecture for a performance management solution becomes a choke point as your monitored domain expands.

The alternative? SevOne provides a unique, patented architecture based on the concepts of distributed computing. The absence of a centralized database allows for unmatched scalability of the monitoring platform.

SevOne deploys as a self-contained physical or virtual appliance. Each appliance acts as the collector, database, and reporting engine. A single SevOne appliance monitors up to 200,000 objects. When you peer many appliances together, there is no theoretical limit to the number of objects SevOne can monitor. How is this possible?

Each appliance maintains two databases: a configuration database and a performance metrics database. SevOne duplicates the configuration database across all appliances in a cluster. Yet, each appliance maintains its own performance metrics database. Users can run global reports from any appliance. The reporting appliance sends parallel requests to its peers and then compiles the results for the report.

Essentially, SevOne maintains speed at scale because it distributes report demands across multiple appliances.

SevOne's Unique Performance Metric Architecture

SevOne’s unique architecture eliminates the concept of a centralized database for performance metrics. This allows for unparalleled speed at scale.

The centralized database architecture typical of many Performance Management solutions inhibits scalability. Some solutions achieve levels of scalability, but only at the expense of speed, money, and time to deploy. Only SevOne allows for unlimited ability to monitor everything in your environment, at speed.

Written by Rick Stuby
Product Marketing Manager

Rick Stuby is a product marketer for SevOne. A strategy, marketing and business development professional with 28 years of global technology experience in networking, telecommunications, semiconductors, renewable energy, and defense covering hardware, software and service solutions, Rick holds eleven United States patents. He has written many technical and marketing publications.

Subscribe To Our Blog