NetCrunch Platform

TOP BENEFITS

1. High performance - if you need to monitor only 50 or 100 devices it works well on low-end hardware (even Intel NUC or Compute Stick can do it), it works well on VM, and you can smoothly scale up to 10,000 devices
2. Universal, feature rich platform regardless of monitoring scope
3. No need for additional software and licenses such as web server or database servers
4. NetCrunch is customizable and lets you access its data in a various way (ODBC, data export, integrations). It's hard to list all things that can be customized in NetCrunch. For example, the console supports docking, multiple screen management, and automatic full-screen display mode.
5. An advanced live presentation including dashboards and graphical maps with widgets. Also, support for Grafana (and Grafcrunch)
6. Central logic makes writing plugins and custom scripts easy - they focus only on delivering state and metrics of monitored objects
7. NetCrunch is watching itself - the key to the NetCrunch reliability and scalability is extensive monitoring of NetCrunch internal processes. Wouldn't you like to know first that monitoring system is overloaded?

The Atlas

The Atlas is a central database containing all your monitoring data. The hierarchy of the Atlas views organizes your data into a manageable database.

In other words, Atlas is all about data organizing. NetCrunch allows creating custom views based on filters and also in a more dynamic way using dynamic folders. Dynamic Folder groups nodes by specific criteria and then each group can be placed on a separate view or put on a single one depending on the number of elements.

Each view represents a group of nodes, and it can be presented in various ways such as a map, dashboard or a table specific to the monitoring data. It's customizable, and program chooses the best tab for the given view.

Customizable

NetCrunch allows you to extend node properties by adding multiple custom fields, for example related to its location details. In this way, you can add any information to the nodes, and you can define reports, policies or filtered views based on custom fields.

Monitoring Dependencies

When you monitor devices behind a router, switch or bridge, you do not want to see the alert storm when the intermediate connection is broken.

Thanks to the dependencies, NetCrunch can disable monitoring of nodes behind the failed device and also can reduce false alerts caused by the link failure.

The top dependency is the NetCrunch machine - if it loses connection, it stops monitoring everything. Additionally, NetCrunch monitors its internal processes to make sure everything is working fine.

Graphical Maps with Widgets

We built maps for presenting network topology and dependencies. After we added live data widgets, they became another type of dashboard where you can present even weather data.

@@img:map-dashboard-weather.png Map with Weather Info

Here is another example of the map created from various widgets, representing an ESX cluster.

@@img:map-dashboard.png Map Dashboard

NetCrunch also creates automatically laid out maps for views that can dynamically change (nodes appear or disappear). This especially applies to Layer 2 segment maps.

Overview Dashboards

NetCrunch offers two views for a status overview. The first one is more detailed, the second one is aggregated.

Nodes Overview

For a smaller number of nodes, it shows nodes detailed view of each service status per node.

Summary Overview

For a larger number of nodes, summary view aggregates statuses in a more concise, per service view.

Top Charts

Another easy to configure dashboard showing top statistics for the Atlas view.

@@img:top-charts.png Top Charts

There are many pre-defined charts included such as

• Top Problem Nodes
• Processor Utilization
• Free Disk Space
• Memory Usage
• Slow Responding Nodes (by PING)
• Network Traffic (Switches and Routers)
• Network Traffic (Servers)
• Top Interfaces by Traffic
• Most loaded nodes by disk read/write
• Network Traffic Summary
• Processes by CPU load
• Processes by Memory
• Top Interfaces % Utilization

You can create a top chart based on any monitoring performance counter.

Performance Views

@@img:performance-view.png Performance View

NetCrunch provides a particular type of Atlas view called Performance view. It can present charts, gauges and bar charts with live data. You can create a view for a group of nodes by defining a filter for nodes and then by adding counters to be displayed (i.e., "% Processor Utilization" for servers), or you can add each chart manually. Additionally, you can group multiple trends on a single chart.

@@img:performance-view-group-chart.png Performance View - Grouped Charts

Performance Monitoring

Monitoring performance metrics is a core of every monitoring system. NetCrunch stores every collected metric in the Performance Trend database - fast, compressed, file-based NoSQL database.

In any place where performance metrics are accessible, you can always look into the historical data of the particular counter, and compare multiple counters' values over time.

@@img:trend-viewer.png NetCrunch Trend Viewer

Calculated Performance Counters

Sometimes we need to monitor the calculated metric based upon other counters presented by the device. The example can be Windows "Normalized Processor Usage" which requires system processor utilization to be divided by the number of processors.

@@img:calculated-counter.png Calculated Counter

GrafCrunch

@@img:grafcrunch-dashboard.png GrafCrunch Dashboard

NetCrunch comes with a fork of the open source project Grafana -one of the top open source performance visualization projects,

it greatly increases the possibilities of creating live performance dashboards and allows you to present data from various sources. You can also put on a single dashboard data from multiple NetCrunch servers.

GrafCrunch has been created for easy installation and integration into NetCrunch.

Event Log

NetCrunch act as a central log server for various event sources. It stores events in the NetCrunch Event Log (SQL) database and performs defined alert actions.

Pending Alerts

In NetCrunch, all internal alerts are automatically correlated, so the program knows when an alert begins (and is pending for resolution) and when it is finished (closed or resolved). NetCrunch allows to easily correlate even external alerts such as traps or syslog messages to provide an accurate state of the system.

Event Details

For every event stored in the event log, NetCrunch offers a detailed view. Beside showing parameters of the event, the window allows you to manage alert settings, and to find where it is defined (in which policy), You can see useful frequency analysis which tells you how often given type of event would happen on given node. Additionally, administrators can add many comments to it.

Alerting Actions

As a response to an alert, NetCrunch can execute a sequence of actions, and they can escalate over time. Operations can also be executed when the alert ends (on alert close).

@@img:alerting-actions.png Alerting Actions

A typical action is a notification, but NetCrunch also allows (remote) executing of corrective actions like commands (i.e. restarting) or scripts.

Common action groups:

• Basic Actions: Such as playing sound, sending desktop notification
• Control Actions: Such as running a program or script
• Logging Actions: You can write events to a text file or to Windows Event Log, send SNMP trap, send syslog message or trigger WebHook
• Scripts: Run remote scripts on Windows or Linux family systems

Read more about Alerting Actions

Reporting

NetCrunch includes comprehensive viewer allowing generating reports on demand and accessing stored reports that can be generated automatically.

All reports can be scheduled and sent as email attachments to administrators. Reports are in PDF format.

@@img:report-viewer.png Report Viewer

Predefined Reports

NetCrunch contains about 50 pre-defined reports. The reports are divided into two categories for Node and for Map as they target detailed information about the node or about the map.

Basic node reports include

• Node Availability Report
• Node Summary Report
• TCP/IP Report
• Node Performance Counters View
• Network Traffic on Node Interfaces
• Node Interfaces Availability
• Node Network Services and Interfaces

Basic map reports

• Detailed Report of SNMP Enabled Nodes
• Details of Node Monitoring Report
• Nodes Properties Report
• Nodes Summary Map Report
• Physical Segments Report
• Top 10 Nodes Availability Report
• Interfaces Traffic Report

Custom Reports

Defining Reports for Atlas Views

You can easily define report based on table view of any of the Atlas View. It can be created from Nodes view or any monitor view such as SNMP or Windows.

This feature can be particularly useful if you define your own custom fields, add them to nodes, and you want a report related to these columns.

@@img:report-from-view.png Report From the View Wizard

Defining Reports on Performance Metrics

It's very easy. You can define Data Collector and add it to Monitoring Pack or to the node.

@@img:data-collectors.png Data Collector Settings

NetCrunch Self Monitor

NetCrunch is a scalable, multi-process and multi-threaded asynchronous system.

It processes millions of operations 24x7, and you can rely on it. It consists of many internal processing elements, tasks, and queues. No wonder we have to measure hundreds of parameters inside of NetCrunch to know when and why it might break.

There is no such thing like infinite scalability (the same as no clean diesel engines) claimed by some German software vendors. The system can always run out of resources.

NetCrunch internal architecture looks more like it is a child of the application server and database server with some genes of the operating system.

We strive to improve NetCrunch performance and reliability with every new version, as we always want it to be the best NetCrunch release ever.