Fetch Cluster ID and Name in Metrics Queries Using DQL

[manuanand]

Hi all,

I am creating a Dynatrace dashboard for which I need data from a particular cluster, I am currently using AWS Container Insight metrics in Dynatrace data explorer, specifically metrics like:

ext:cloud.aws.containerinsights.pod_cpu_reserved_capacityByNamespacePodName:splitBy(Namespace):sort(value(auto,descending)):limit(20)

My goal is to filter these metrics to show data only from one specific cluster. However, when I try using filter expressions like:
:filter(eq("ClusterName","my-cluster")), i get the error :
The Dimension key 'ClusterName ' (or 'clusterId') has been referenced, but the metrics have no such key.

I have also inspected the available dimensions for this metric using splitBy(), and I don't see any cluster-level dimensions exposed.

My Questions:

1. Is there a version of these Container Insights metrics that includes a cluster name or similar dimension?

2. If not, how can I reliably filter these metrics per cluster?

I'd appreciate any help or examples from those who've dealt with this limitation. Also, I want to alias this metric, but I am not able to figure it out.

Thanks!

[marco_irmer]

I noticed you posted this thread in the DQL section of the forum.. Are you open to using a DQL query in a Notebook or Dashboard for this task? The reason I ask is because DQL offers additional flexibility compared to metric selector expressions, enabling multi-step queries to add fields (including entity attributes and relationships) or conduct further filtering.

[manuanand]

Yes, i am writing the metric dql in Data Explorer of Dynatrace dashboard, i have limitation to use only the same. Please check the screenshot for clarity. If anything can be done to alias the metric expression and filter the data at cluster level in a single statement, that is needed or if any alternate approach to get expected results.

[IzabelaRokita]

Hey @manuanand ,
I just wanted to check in and see if you still need help with this. If so, I’d be happy to look into it for you! 😊
Please let me know what works best for you.

[manuanand]

Yes please, that would be helpful.