opentelemetry-demo icon indicating copy to clipboard operation
opentelemetry-demo copied to clipboard

Published 20 hours ago • verified publisher icon open-telemetry

No graphs for "Diagnosing memory leaks" scenario

Open Kielek opened this issue 1 year ago • 4 comments

Bug Report

Which version of the demo you are using? (please provide either a specific commit hash or a specific release).

Current main 02837fe7b71dd3b85db0efb253a97a693b5d5e3a - it downloads 1.4.0-beta images.

Symptom

What is the expected behavior?

Data similar to https://opentelemetry.io/docs/demo/scenarios/recommendation-cache/

What is the actual behavior?

Following charts are not displayed under demo dashboard http://localhost:8080/grafana/dashboards image (3)

Recommendation Service (CPU% and Memory) Service Latency (from SpanMetrics) Error Rate

Reproduce

  1. Download repository
  2. Execute docker compose up (note it will download
  3. Enable recommendationCache feature flag
  4. Wait some time (tested 1h+)

Additional Context

Tested on Windows 10 with Docker with Linux containers.

Kielek avatar Apr 13 '23 10:04 Kielek

Most of graphs in "OpenTelemetry Collector" dashboard also has no data.

pellared avatar Apr 14 '23 11:04 pellared

Hi, Mr. Kielek I met the same problem. I wonder if you resolve this problem?

LyndonZhao avatar May 27 '23 02:05 LyndonZhao

I have a similar problem in that I don't have any data in the recommendation graphs. I noticed that the metric runtime_cpython_cpu_time is not available in Prometheus. I only have runtime_cpython_cpu_time_seconds_total.

dbcummings3174 avatar May 30 '23 17:05 dbcummings3174

It might be just a metric name mismatch in Grafana vs Prometheus. For example for the latency graph click on its title, edit, and replace the metric name duration_milliseconds_bucket with duration_bucket in all four queries. The error rate one might need to be updated to calls instead of calls_total.

Notice also that the $service variable might need to be updated (go to dashboard setup > variables). If you update this variable definition to 'label_values(service_name)' the 'service' drop down menu on top right of the dashboard will show all services for selection (it won't work for all graphs though e.g. the Recommendations Count graph will obviously only work with the recommendation service selected at the drop down.

mrcazzim avatar Aug 15 '23 16:08 mrcazzim

Closing this as part of issue cleanup. If there's anything that needs to be done, please create a new issue.

austinlparker avatar Feb 19 '24 16:02 austinlparker

@austinlparker Can you please reopen it? I do not look that the bugs have been addressed.

pellared avatar Feb 19 '24 16:02 pellared

Can you provide a repro case for 1.8? I feel like this was fixed.

austinlparker avatar Feb 19 '24 16:02 austinlparker

Can you provide a repro case for 1.8? I feel like this was fixed.

Please assign it to me. I will take a look tomorrow or on Wednesday if the issue still exists.

pellared avatar Feb 19 '24 16:02 pellared

@austinlparker It is fixed. Feel free to close.

pellared avatar Feb 20 '24 15:02 pellared

@pellared the problem still exists.

I have run a helm repo upgrade and then helm upgrade <release>

Then when I run helm history <release> the app version is still the same.

REVISION UPDATED STATUS CHART APP VERSION DESCRIPTION
1 Thu Feb 29 10:25:31 2024 superseded opentelemetry-demo-0.29.0 1.8.0 Install complete 2 Sun Mar 10 18:40:00 2024 deployed opentelemetry-demo-0.29.0 1.8.0 Upgrade complete

Can this issue please be reopened and looked into for resolution.

jas-atwal avatar Mar 10 '24 19:03 jas-atwal