Page History

NOTE - With the exception of the thresholds/exceptions the values used below are configurable by the user.

 This document outlines the various report types that opReports offers.

Node Report

The Node Report provides details about a selected a detailed summary of  one node. Node details such as status, sysName, ip address, type, model, uptime, interfaces, location, contact, description, last update, vendor, object name, group role and net and the interface table are providedpresented.

Interfaces and storage items are details if present.

Graphs are provided which details the following: reachability, availability and health, response time, CPU utilisation, number of routes, ip utilisation, IP fragmentation/reassembly (as a % of packets received), buffer utilisation.

Items of note:

• This report relies directly on NMIS  for inline graphs, and therefore won't work unless the configuration item nmis_host_base is correctly set (i.e.

...

• has the public web address of your NMIS server).
• Business Hours reporting is not supported for this report.
• This report cannot be generated in formats other than HTML.
• While it can be saved it is not self-contained (the NMIS graphs are live and created on viewing!), therefore it's primarily of use as an on-demand report for immediate consumption.

...

Node Health Report

Display the The Node Health report display health-related attributes for a all selected node nodes for a given period. Attributes displayed are: Status, Device, Availability, Interface Availability, %CPU, 95th% CPU, Max %CPU, CPU Exc., %Mem Free, 95th% Mem Used, Max %Mem Used, %Mem Util, %IO/VIR Mem Free, 95th% IO Mem Used, Max %IO Mem Used, %IO/VIR Mem Util, Conditions, Actions.Interface charts are provided for packets per second, errors and discards, bits per second and CBQoS for each interface.

The report also includes two columns with the detected (abnormal) Conditions and the recommended Actions.

If you pass this report the option exceptions=true, then only nodes with exceptional conditions present are shown; the default is to show all nodes.

The formulas used for calculation of the reporting conditions are configured in the opReports Configuration file.can be tuned and adjusted by the user:

The section opreports_rules (in conf/opCommon.nmis in opReports 3.x, or opReports.nmis in version 2.x) defines the threshold values for the following conditions:

Device Availability = Condition: "Device has LOW or VERY LOW availability"
Action: Investigate causes for low availability
Formula used for Calculation:

...

• Low free main memory less than 25

WAN Report

Display The WAN report displays the WAN Link performance for a selected nodenodes.

Node details displayed are: status, conditions, actions, device, availability and response time.

For each interface on the node, the following are displayed: interface, speed, average utilisation, maximum utilisation, average inbound errors (in %), average outbound errors (in %), average inbound discards (in %), average outbound discards (in %), average inbound utilisation (in %), average outbound utilisation (in %), maximum inbound utilisation (in %), maximum outbound utilisation (in %).

The WAN report health rules are configurable (section opreport_rules) and the report type supports customisable detail levels for the display.

Formula used for Calculation
Maximum and Average input and output utilisation is calculated for the interface. The larger of maximum and average input and output utilisation is used in formula.

...

The QoS report is intended to provide an overview of the QoS Qality of Service configuration and utilisation of the network. NB. To have QoS reporting, NMIS has to be collecting CBQoS statistics, which is configured in the NMIS System Configuration, updating the Node details.selected nodes.

Rules
To assist in the assimilation of the data, a few simple rules have been applied to the statistics to assist with highlighting some possible problem areas.

...

Class Utilisation Exceeding Policy
This condition indicates that QoS class was configured for a certain bandwidth and the traffic levels have exceeded the configured policy. This may not be a problem, depending on QoS configuration, traffic may be being forwarded or discarded, if necessary increase the bandwidth allocated to this class.
  If an interface utilisation is greater than 0% input and 0% output;
  AND PrePolicyUtil exceeds the classes configured bandwidth Percent;
  THEN this class is highlighted yellow. 

Uptime Report

Display the uptime data for a given node.

The Uptime report is intended to provide The uptime repors provides an overview of devices which have restarted in the last 7 days as well as devices which have been up for over 365 days.

A reason for the devices restarting in the last 7 days should be known.

recently restarted devices, as well as very long running ones.

The configuration items uptime_shortest_days (default 7) and uptime_longest_days (default:365) define which nodes should be selected for display. 

Response Time Report

The Response Time report tabulates the selected nodes in descending order of their average response time. Besides the average and maximum measurements, the report also shows the 95th percentile of the response time. These readings are in milliseconds.

...

This column shows the number of times a device's response time has exceeded the exception threshold during the reporting period. This threshold is configured in the opReports.nmis configuration fileconfigurable (property response_exception_threshold), and its current value is 10 msdefault value is 10 ms.

Interface Utilisation Report

The interface utilisation report shows the interface utilisation statistics for one or more interfaces. By default it will display the averages for input, output, combined and higher-of-in-and-output bandwidth utilisation, as well as exception counts and cumulative exception period.

Exceptions are defined as any of the utilisation readings rising above option util_threshold (default: 80%). The exception period is defined as all the intervals with over-threshold readings. In addition to those raw readings, the labelling of an interface as in exceptional or normal state is controlled by option util_threshold_mincount (default: 1), which defines how many exceptions have to be observed before the interface is labelled "bad".

This report can be further adjusted with these options:

• Option show_threshold (default: true)
  If set to false, no thresholds are shown; instead the bandwidth, average traffic and average utilisation are presented (plus a shortened report period column).
• Option show_only_util (default: false)
  If show_treshold is false, and if show_only_util is set to true, then only bandwidth and average utilisation are shown (ie. average traffic is omitted).