NMIS

History

NMIS started as an SNMP polling and statistics viewer front-end to Tobias Oetiker’s RRDTool. At the time RRDTool and MRTG did not include a front or back end to handle SNMP polling and to display resulting information in a web interface. NMIS was written to fill in the gaps in existing available products and provide one system which met the large number of requirements of network managers. NMIS was developed to be proactive, provide reporting, and give the status of the network "at a glance" by Keith Sinclair in 1998. Mr Sinclair released the code as open source and nurtured community involvement.

Eric Greenwood become a major contributor in 2004 and is acknowledged as a co-founder.

In 2011 the intellectual property rights to NMIS were acquired by Opmantek. Opmantek has added significant development to the free NMIS product while creating numerous commercial modules which they license for a fee.

Overview

NMIS performs multiple network management functions from the OSI Model and International Organization for Standardization FCAPS model, these being - Fault, Configuration, Accounting and/or Administration, and Performance. These metrics provide valuable capabilities and features for fault and performance management, which in turn are useful for many other aspects of network and business management. NMIS monitors the status and performance of an organization’s IT environment, assists in rectification and identification of faults and provides valuable information for IT departments to plan expenditure and IT changes.

The NMIS business rules engine classifies events on their business impact, not just the technical nature. The rules engine is extremely powerful; however it can be configured in minutes for a network with a small number of devices to hours for networks with large numbers of devices.

NMIS uses a single poll for performance and fault data, which reduces the bandwidth of the network management traffic and increase the performance of the network management system. The returning data creates real-time performance monitoring and graphing. When NMIS probes are deployed throughout the network, the network can be managed easily to avoid bottlenecks and enable zero cost redundancy. Both the front and back ends of NMIS are highly extensible and features are easy to add. Custom statics can be gathered for any metric available on a device.

NMIS is very popular within Telecommunications carrier organizations and Managed Service Providers (MSP’s) and is used by IT staff as a key business management and improvement tool.

Features

Hardware Requirements

NMIS can run in both physical and virtual environments. Smaller networks can be monitored using workstation PCs. Server hardware with 2GB of RAM is usually sufficient for up to 1000 managed devices (even with 32 ports on each). 1000 managed devices may take 20GB of storage. Large numbers of devices can be managed by implementing NMIS on more powerful machines or by leveraging NMIS distributed polling and setting up multiple NMIS pollers in a master/slave design. Ultimately, NMIS is very efficient but the hardware specification of the required server will vary depending on what is being managed, how often information is collected, and how many devices are managed.

Recommended Requirements:

Software Requirement

NMIS has been developed in Perl and runs natively on Linux, performing best as a 64bit multithreaded application. It is commonly used on 64bit Red Hat Enterprise Linux, or other F/OSS equivalents such as CentOS or Debian. There is also notable activity with NMIS on Ubuntu and BSD. The software is available online as source code (with installer script) or as a Virtual appliance. The Virtual Appliance is compiled as a .OVF image and needs to be converted to a read/write image to store data. It comes with NMIS, CentOS and Apache installed and requires no additional configuration.