chore: add cluster, nodes that are sending incorrect host metrics

[srikanthccv]

Summary

Part of #6890

In our old charts, we were sending the metrics of the container where the agent is running instead of the host machine it is running. We ask users to upgrade to get the correct results. However, that info is not very useful if they upgraded in some clusters but not in others. This change adds cluster and node names that are sending the incorrect host metrics so they can know which cluster needs to be updated.

---

Important

Add functionality to identify clusters and nodes sending incorrect host metrics by updating IsSendingK8SAgentMetrics and HostListResponse.

• Behavior:
  • IsSendingK8SAgentMetrics in hosts.go now returns lists of cluster and node names sending incorrect host metrics.
  • Updates GetHostList in hosts.go to populate ClusterNames and NodeNames in the response.
• Models:
  • Adds ClusterNames and NodeNames fields to HostListResponse in infra.go.

^{This description was created by for 5f3524e. It will automatically update as commits are pushed.}

[ellipsis-dev]

❌ Changes requested. Reviewed everything up to 5f3524e in 1 minute and 20 seconds

More details

• Looked at 98 lines of code in 2 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. pkg/query-service/app/inframetrics/hosts.go:336

• Draft comment:
  Avoid adding non-ClickHouse related logic to the ClickHouseReader interface. Use the DAO in the telemetry instance for such logic instead. This is also applicable to the DidSendHostMetricsData function.
• Reason this comment was not posted:
  Decided after close inspection that this draft comment was likely wrong and/or not actionable:
  The functions in question are actually doing ClickHouse-specific queries - they're querying metric tables with ClickHouse-specific SQL syntax. There isn't really any non-ClickHouse logic here that could be moved to a DAO. The functions are appropriately placed in this repo since they deal directly with querying ClickHouse for metrics data.
  Could there be some business logic mixed in with the ClickHouse queries that I'm missing? The functions do some data processing after getting results.
  The post-processing is minimal and directly related to the ClickHouse query results - just transforming the raw data into the expected return format. This is typical for a repository layer.
  The comment should be deleted because it incorrectly suggests moving ClickHouse-specific query logic out of the ClickHouseReader interface.

Workflow ID: wflow_pjwrefrk46Ec6Hx4

---

Want Ellipsis to fix these issues? Tag @ellipsis-dev in a comment. You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.