Predictive baseline monitoring

[ranjitw]

Hi All,

Does Nagios XI have plugin where it analysis historic data and generate alarm based on the trends ?

Example:
1) The used memory of a webapplication increases from day to day

2) Usually we specify alert range for CPU Utilization in one direction UP . What if the CPU falls to 10% utilization or even lower to 1% utilization? If this is a production server it’s CPU should always be busy processing the business service i.e if its underutilized, that might mean there is a application or database issue which needs attention, just as much as if it is over utilized.

Is there a way we can alert for an upper and a lower baseline deviation, or abnormality in Nagios XI ?

Regards,
Ranjit W

[jdalrymple]

With 2 separate services with separate thresholds, or with a wrapper. What you're seeking (anomaly based threshold monitoring) isn't built into the Nagios code.

[ranjitw]

Hi ,

What would be the approach to generate an alarm based on the trends ?

Regards,
Ranjit W

[jdalrymple]

Again - a wrapper.

I've created plugins personally that alert based upon changes instead of absolute values. In particular temperature changes over short periods of time are alarming.

As I mentioned, the feature you're seeking isn't written into the nagios core monitoring code - you'd have to make your plugin intelligent enough to handle it. If you can be more elaborate on what trends you're seeking to follow I maybe could offer you more advice. I can tell you for certain though you're likely facing a customized plugin.

[ranjitw]

Hi jdalrymple,

Is there a 3rd party plugin which supports this functionality ?

I would post list of features which are needed w.r.to baselining by this week .

Regards,
Ranjit W

[tmcdonald]

I wrote an article a while ago about bischeck, a third-party addon that does what you are looking for:

https://assets.nagios.com/downloads/nag ... ios-XI.pdf

[ranjitw]

Thanks tmcdonald,

I will explore more on the plugin .

Does this plugin support this scenario ? - Every weekend we are having a scheduled backup and during this period the CPU utilization is increasing and alarm is being triggered . Is it possbile to suppress this kind of alerts by seeing the trends ?

Regards,
Ranjit W

[scottwilkerson]

Every weekend we are having a scheduled backup and during this period the CPU utilization is increasing and alarm is being triggered . Is it possbile to suppress this kind of alerts by seeing the trends ?

A possible solutions would be to use a custom time period that doesn't contain the time frame that the scheduled backups run.

If the backups are at a know time, you could also create a recurring downtime schedulle for the triggered CPU that eliminates alerts durring the period that the scheduled backups are being performed.

[ranjitw]

Hi,

Is there a plugin which analysis the trend of the usage of service in host and automatically ignore the alert based on trend?

Example : We have random servers on which backup activity takes place and the CPU usage increases(One scenario) - Is there plugin which automatically detect the patterns and stop alerting ?

Regards,
Ranjit W

[ranjitw]

Hi ,

We want to implement baselining similar to what is mentioned in below URL :

https://communities.bmc.com/servlet/Jiv ... lities.pdf

Is nagios xi having these features or any plugin or 3rd party tool ?