Your metric query has been transformed into a query that is compatible with the previous infrastructure metric format -- what does this mean?

Support: Telemetry Data Platform (TDP) Dashboards
#1

On my dashboard I see a little warning sign (triangle with exclamation mark, :warning:) next to the name of some graphs. The tooltip reads “your metric query has been transformed into a query that is compatible with the previous infrastructure metric format”.

What does this mean? Should I do something, like change the format of the metrics I ingest into New Relic, or the query for my graphs?

#2

Hi there @dschridde -

Thank you for bringing this to the community. There could be a few reasons that this is happening. Could you please take a look at the known limitations listed here:

https://docs.newrelic.com/docs/query-your-data/nrql-new-relic-query-language/nrql-query-tutorials/query-infrastructure-dimensional-metrics-nrql#known-limitations

If you do not believe that your queries are bumping up against any of these limitations, we can dig deeper with you!

#3

Thank you!

Sadly cannot get the :warning: to go away even after reading those instructions.

Originally I had:

FROM Metric SELECT max(k8s.container.cpuUsedCores) AS 'Used', min(k8s.container.cpuRequestedCores) AS 'Requested', min(k8s.container.cpuLimitCores) AS 'Limit' WHERE k8s.namespaceName='...' AND k8s.containerName='...' AND tags.app='...' TIMESERIES

I read in the documentation you pointed out:

In order to select attributes starting with tags. a metric name has to be provided. For example, SELECT uniques(tags.environment) FROM Metric WHERE metricName=’aws.lambda.function.duration’ does not work without the WHERE clause.

Hence I turned this into:

FROM Metric SELECT max(k8s.container.cpuUsedCores) AS 'Used', min(k8s.container.cpuRequestedCores) AS 'Requested', min(k8s.container.cpuLimitCores) AS 'Limit' WHERE metricName IN ('k8s.container.cpuUsedCores', 'k8s.container.cpuRequestedCores', 'k8s.container.cpuLimitCores') AND k8s.namespaceName='...' AND k8s.containerName='...' AND tags.app='...' TIMESERIES

But that still shows the same :warning: and tooltip.

#4

@dschridde Following up to see if you are still seeing the warning sign. Sorry that our support team hasn’t been able to look into this further for you yet.

#5

@JoiConverse, thank you for your response. I still see the same warning symbol for charts using data ingested from Kubernetes.

#6

@dschridde - The warning symbol is referring to the following limitation in the document that @hross posted above:

  • Metric queries with * do not return Infrastructure sample data (for example, SELECT * FROM Metric).

Also, please note that Infrastructure dimensional metrics with NRQL feature is still in beta.

2 Likes
#7

Sadly the :warning: symbol is displayed despite my query not containing *:

FROM Metric SELECT max(k8s.container.cpuUsedCores) AS 'Used', min(k8s.container.cpuRequestedCores) AS 'Requested', min(k8s.container.cpuLimitCores) AS 'Limit' WHERE k8s.clusterName='MYCLUSTER' AND k8s.namespaceName='MYNAMESPACE' AND k8s.deploymentName='MYDEPLOYMENT' AND k8s.containerName='MYCONTAINER' TIMESERIES
#8

Hi @dschridde - did you resolve this issue?

I’m now seeing similar on unchanged dashboards (but thankfully still work). I get a little grey circle containing an “i”. More commonly on AWS integration metrics than Infrastructure or other types.

Here’s the simplest generic example I’m using that gets this “error”;

SELECT sum(`provider.requestCount.Sum`) from LoadBalancerSample WHERE provider = 'Alb' until 10 minutes ago TIMESERIES FACET displayName

I would love to know what the “new Metric format” is so that this could be rectified. :slight_smile:

It’s not so bad for in house dashboards, but when creating for a client they invariably ask “what does that mean?”. And right now I don’t have a good answer. :frowning:

#9

Hi @dkoyano, I’m also having a similar issue but my query doesn’t seem to fall under any of the limitations category. This is the query I’m using:

SELECT min(provider.FreeStorageSpace.Minimum / (provider.ClusterUsedSpace.Minimum + provider.FreeStorageSpace.Minimum)) * 100 as ‘Free storage space (%)’ FROM DatastoreSample WHERE provider=‘ElasticsearchCluster’

Is there something incorrect about doing the arithmetic in this way?