Insights Explorer says No Data to Display in anomaly model testing

[Karthick Nagara]

Hi All,

Im currently trying to get the metric data from Zabbix to Servicnow operational intelligence. Im able to see the Metric to CI data for the Windows server im monitoring.

Since I got only 8 out of 103 metric types got mapped to the WIndows server I added an Event Rule to map the host to maximum metrics as mentioned in the URL, Create Event Rule to Bind Metric Event to host Cis

Im trying out Operational intelligence for the first time and I want to see how the Anomalies would look like which is the reason I created a Event rule to map most metrics to the Windows server.

After doing so , I created an Anomaly model Testing For only Tomcat related metrics.and saw the results having no scores and even the insights explorer says No data to display.

I even tried restarting the Tomcat service in my windows server being monitored but Im not able to see the metric data , even the Grey Dots that would indicate the metrics being tracked.

 

This is the case for any metric (not only Tomcat) . I want to see atleast one metric/anomaly to analyse further on this with respect to the server im monitoring.

Can any one help me if im missing any steps in the configuration of Operational Intelligence...

Please Help.

[Gianpaolo Pagan · Accepted solution]

Hi Karthick,

Are you actually seeing any metric in "Insights Explorer"?

Operational Intelligence has a dependency on MetricBase, a time-series database and the related plugin (com.snc.clotho) needs to be enabled before you can use Operational Intelligence.

If the plugin is enabled, I suggest you to

1. go to sys_clotho_config.list
2. There should be only one record with name clotho
3. If you have that record run the UI action "test connection" and see if the tests succeed

Please let me know if that was the case otherwise we can look further.

I hope this helps,

Gp

[Karthick Nagara]

Thanks Gianpaolo.

I tested the connection. It was failing in my Personal development instance.

Could you please help me understand what configuration need to be done here.

And I hope Metric base is the engine that does the Statistical model and detect anomalies kind of stuff. Correct me if im wrong.

[Gianpaolo Pagan]

MetricBase is the TimeSeries database. It's where the metrics, baselines and anomalies are stored. As a timeseries database, MetricBase allows to query timeseries data selecting various types of aggregations and manipulations (avg, min, max, deviations, counts, median and many others), it also allows touse triggers on metrics to start flows in FlowDesigner/IntegrationHUB and execute actions as a result of that trigger.

Even though MetricBase is not what actually performs the learning and anomaly detection, these things are not possible without it.

I don't know if things work differently with dev instances but the normal procedure to get MetricBase provisioned and connected to an instance is by requesting the plugin enablement on the Hi portal.

Many aspect of OI are driven by jobs. Learning is done by a job called "Operational Intelligence - Metric Learner job". Anomaly detection happens on the MID server so that detected anomalies can be quickly sent to the instance without having to wait for the metrics to be shipped and stored.

Gp

[Karthick Nagara]

Than you.

I tried the same with another instance (Not a PDI). There  test connection was success ful in sys_clotho_config.list . Also im able to see the Raw metrics graph lines in Insight explorer as well.

Not sure why it didnt work in personal developer instance.

One more clarification please, does operational intelligence maintain a database in MID server?

Through Mertricbase I see all metric columns are maintained in same Servicenow database . But the thing that confused me is the below link where it is mentioned that there is a property

connection_login_timeout_secs
to

Maximum time in seconds to log in to the local database on MID Server.

https://docs.servicenow.com/bundle/london-it-operations-management/page/product/event-management/task/create-config-overriding-rule.html