cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Looking to upgrade from Dynatrace Managed to SaaS? See how

Metric Ingestion - Query

Hasan
Dynatrace Enthusiast
Dynatrace Enthusiast

Hi There,

Using metric ingest we have created a metric which is configured for 2 servers (1 live server and 1 is fallback server ), the client wants alert to be generated only when both the nodes are not fulfilling the requests, refer screen below for a sample metric created, kindly advise how we can fulfil this.
Please note: with the current configured metric we are getting alerts for both the nodes as separate alert.

 

Previously posted this in slack so far no reply 

https://dynatrace.slack.com/archives/C01P9N2RZQE/p1693290399106039

 

3 REPLIES 3

mark_bley
Dynatrace Champion
Dynatrace Champion

Hi @Hasan ,

I am assuming both nodes are will send the same metric key.

When both of them fail no metric data points will be sent into Dynatrace so data will be missing

-> "alertOnNoData": true,

Using a metric event like following would alert you if the metric is completely missing, i.e. primary and secondary both have failed.

 

[{
    "schemaId": "builtin:anomaly-detection.metric-events",
    "schemaVersion": "1.0.15",
    "scope": "environment",
    "value": {
      "enabled": true,
      "summary": "Important Metric is missing",
      "queryDefinition": {
        "type": "METRIC_SELECTOR",
        "metricSelector": "very.important.metric.that.has.to.be.there",
        "managementZone": null,
        "queryOffset": null
      },
      "modelProperties": {
        "type": "STATIC_THRESHOLD",
        "threshold": -1,
        "alertOnNoData": true,
        "alertCondition": "BELOW",
        "violatingSamples": 3,
        "samples": 5,
        "dealertingSamples": 5
      },
      "eventTemplate": {
        "title": "Very important metric is missing",
        "description": "The {metricname} value was {alert_condition} normal behavior.",
        "eventType": "AVAILABILITY",
        "davisMerge": true,
        "metadata": []
      },
      "eventEntityDimensionKey": null,
      "legacyId": null
    }
  }
]

 

Hasan
Dynatrace Enthusiast
Dynatrace Enthusiast

Hi Mark,
Both nodes are sending separate metric key sharing the same below along with screenshots.
Metric in Live server

icardonline.prod.host119.FALCON.55382

Metric in Fallback server

icardonline.prodfb.host118.FALCON.55382

 

mark_bley
Dynatrace Champion
Dynatrace Champion

Hi @Hasan 

I cannot recommend using dimensions in the metric key, that is what metric dimensions are for.

https://www.dynatrace.com/support/help/shortlink/metric-ingestion-protocol

You would send a metric line like this and both hosts would send this with different dimension values e.g.:

icardonline,stage=prodfb,host=host118.FALCON.5538 <metric data>
icardonline,stage=prod,host=host119.FALCON.55382 <metric data>

 

If you do not want or cannot use above's format, you still could use a metric query in the metric event adding up both metrics and alerting if data is missing, but again this is not recommended.

 

BR,

Mark

 

Featured Posts