[O11y][Prometheus Input] Add system test cases #6490
Conversation
💚 Build Succeeded
Expand to view the summary
Build stats
Test stats 🧪
🤖 GitHub commentsExpand to view the GitHub comments
To re-run your PR in the CI, just comment with:
|
| @@ -0,0 +1,4 @@ | |||
| vars: | |||
| hosts: | |||
| - http://{{Hostname}}:{{Port}} | |||
There was a problem hiding this comment.
We are not collecting data of a specific service having a Prometheus end point here ?
There was a problem hiding this comment.
@ishleenk17 Do you mean like setting up any service such as WebSphere Application Server, IBM MQ, etc. and then collect the metrics using Prometheus?
| "job": "prometheus", | ||
| "slice": "queue_time" | ||
| }, | ||
| "prometheus_engine_query_duration_seconds_count": { |
There was a problem hiding this comment.
Why only two metrics show up the sample event? Does not prometheues :9090/metrics has many more metrics?
There was a problem hiding this comment.
@lalit-satapathy : That's why I have asked @aliabbas-elastic to have a service running and then run system test. Otherwise by default if we just bringup a Prometheus Input, we get these default metrics like job, labels only.
There was a problem hiding this comment.
Assuming this runs the prom/prometheus image and loads the metrics from :9090/metrics which is fine IMO. I don't know why all other metrics are not showing up in sample event.
Example:
prometheus_target*_ metrics
There was a problem hiding this comment.
Why only two metrics show up the sample event? Does not prometheues :9090/metrics has many more metrics?
Actually I am getting all the metrics but they are in different response. Prometheus logically groups the metrics under different labels and Prometheus Input ingests these groups in independent documents. But in sample_event.json only one document can be displayed.
There was a problem hiding this comment.
Yes looks like prometheus package is creating separate event for each of the Prometheus metric, looks like 1 metric per document.
There was a problem hiding this comment.
@lalit-satapathy @aliabbas-elastic
I don't think it's 1 metric per document.
Please refer to the screenshot in this PR.
Here I have run Prometheus input for the Influxdb service.
There was a problem hiding this comment.
strange, I saw single metric per document in my run.
There was a problem hiding this comment.
Even in InfluxDB it sometimes gives single metric per document. So I think it is grouping that happens based on the quantile label provided go_gc_duration_seconds{quantile="0.75"} 0.000114446
There was a problem hiding this comment.
So looks like grouping is happening by labels.
How about the metrics which don't have labels. How do those show up ?
There was a problem hiding this comment.
Those will be received in one single document under two labels prometheus.labels.instance and prometheus.labels.job. Please refer to the sample_event.json file.
… into prometheus_input_system_test
|
Is there any risk in the linked docker image going away and breaking our tests in future? otherwise LGTM. |
No I don't think so there will be any risk as such as it an official image of the WebSphere Application Server. |
|
Package prometheus_input - 0.1.3 containing this change is available at https://epr.elastic.co/search?package=prometheus_input |
What does this PR do?
Checklist
changelog.ymlfile.How to test this PR locally
Related issues