opentelemetry-instrument 0.53b0 release breaks azure-monitor-opentelemetry

[cforce]

2025-04-12T10:04:15.99111 Connecting to the container 'main'...
2025-04-12T10:04:16.03838 Successfully Connected to container: 'main' [Revision: 'capps-backend-jfnrv5kcsyxqa--azd-1744450658-dbdb5c7d7-r5q9v', Replica: 'capps-backend-jfnrv5kcsyxqa--azd-1744450658']
2025-04-12T09:57:07.4151555Z stderr F 2025-04-12 09:57:07,364 - WARNING - Exception occurred when instrumenting: flask.
2025-04-12T09:57:07.4151567Z stderr F Traceback (most recent call last):
2025-04-12T09:57:07.4151589Z stderr F File "/usr/local/lib/python3.11/site-packages/azure/monitor/opentelemetry/_configure.py", line 200, in _setup_instrumentations
2025-04-12T09:57:07.4151601Z stderr F conflict = get_dist_dependency_conflicts(entry_point.dist) # type: ignore
2025-04-12T09:57:07.4151613Z stderr F ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-04-12T09:57:07.4151625Z stderr F File "/usr/local/lib/python3.11/site-packages/opentelemetry/instrumentation/dependencies.py", line 51, in get_dist_dependency_conflicts
2025-04-12T09:57:07.4151638Z stderr F for dep in dist.requires: TypeError: 'method' object is not iterable TypeError: 'method' object is not iterable
2025-04-12T09:57:07.4151650Z stderr F TypeError: 'method' object is not iterable
2025-04-12T09:57:07.4151662Z stderr F 2025-04-12 09:57:07,372 - WARNING - This is a warning message
2025-04-12T09:57:07.4151674Z stderr F 2025-04-12 09:57:07,375 - INFO - Started server process [13]
2025-04-12T09:57:07.4151686Z stderr F 2025-04-12 09:57:07,375 - INFO - Waiting for application startup.
2025-04-12T09:57:08.3320300Z stderr F 2025-04-12 09:57:08,331 - INFO - Application startup complete.
2025-04-12T09:57:08.3627583Z stderr F 2025-04-12 09:57:08,362 - INFO - Application startup complete.
2025-04-12T09:57:08.3641281Z stderr F 2025-04-12 09:57:08,363 - INFO - Application startup complete.
2025-04-12T09:57:08.3847667Z stderr F 2025-04-12 09:57:08,384 - INFO - Application startup complete.
2025-04-12T09:57:08.4119500Z stderr F 2025-04-12 09:57:08,411 - INFO - Application startup complete.
2025-04-12T09:57:08.4410831Z stderr F 2025-04-12 09:57:08,440 - INFO - Application startup complete.
2025-04-12T09:57:08.4437162Z stderr F 2025-04-12 09:57:08,443 - INFO - Application startup complete.
2025-04-12T09:57:08.4463553Z stderr F 2025-04-12 09:57:08,446 - INFO - Application startup complete.
2025-04-12T09:57:08.4470302Z stderr F 2025-04-12 09:57:08,446 - INFO - Application startup complete.

---

There is a known issue i found
open-telemetry/opentelemetry-python-contrib#3415

---

[cforce]

Its a rabbit whole. I guess i have got over the initial issue by pinning versions in
requirements.in like below....

azure-monitor-opentelemetry==1.6.7
opentelemetry-instrumentation==0.53b0
opentelemetry-instrumentation-asgi==0.53b0
opentelemetry-instrumentation-httpx==0.53b0
opentelemetry-instrumentation-aiohttp-client==0.53b0
opentelemetry-instrumentation-openai

... but now i have a different error message..

Connecting...
2025-04-13T09:12:25.94145  Connecting to the container 'main'...
2025-04-13T09:12:25.99666  Successfully Connected to container: 'main' [Revision: 'capps-backend-jfnrv5kcsyxqa--sxq76iz-b54d6cdc4-2xh9g', Replica: 'capps-backend-jfnrv5kcsyxqa--sxq76iz']
2025-04-13T09:09:27.4808462Z stderr F   File "<frozen importlib._bootstrap_external>", line 940, in exec_module
2025-04-13T09:09:27.4808474Z stderr F   File "<frozen importlib._bootstrap>", line 241, in _call_with_frames_removed
2025-04-13T09:09:27.4808486Z stderr F   File "/usr/local/lib/python3.11/site-packages/opentelemetry/instrumentation/django/__init__.py", line 244, in <module>
2025-04-13T09:09:27.4808550Z stderr F     from django import VERSION as django_version
2025-04-13T09:09:27.4808567Z stderr F ModuleNotFoundError: No module named 'django'
2025-04-13T09:09:27.4808581Z stderr F 2025-04-13 09:09:26,877 - WARNING - This is a warning message
2025-04-13T09:09:27.4808594Z stderr F 2025-04-13 09:09:26,881 - INFO - Started server process [17]
2025-04-13T09:09:27.4808613Z stderr F 2025-04-13 09:09:26,881 - INFO - Waiting for application startup.
2025-04-13T09:09:27.4808626Z stderr F 2025-04-13 09:09:26,883 - WARNING - This is a warning message
2025-04-13T09:09:27.4808664Z stderr F 2025-04-13 09:09:26,888 - INFO - Started server process [13]
2025-04-13T09:09:27.4808678Z stderr F 2025-04-13 09:09:26,889 - INFO - Waiting for application startup.

Microsoft bundles Django, Flask, FastAPI, etc. instrumentation into azure-monitor-opentelemetry, even if you're not using them.
That's why opentelemetry-instrumentation-django shows up - but it shall not try to force instrumentation when Django is not present - i this a bug in opentelemetry-instrumentation auto detection?

I try below and deploy again to azure aca (only on azure execution it fails, not locally)

export OTEL_PYTHON_DISABLED_INSTRUMENTATIONS=django 
azd env set OTEL_PYTHON_DISABLED_INSTRUMENTATIONS django

https://opentelemetry.io/docs/zero-code/python/configuration/#disabling-specific-instrumentations

[cforce]

Solved (added it even as even env var for the container in the bicep

Next rabbit 🐰 😭

2025-04-13T10:06:29.37506  Connecting to the container 'main'...
2025-04-13T10:06:29.48061  Successfully Connected to container: 'main' [Revision: 'capps-backend-jfnrv5kcsyxqa--azd-1744537379-6f6f8777b5-czhmq', Replica: 'capps-backend-jfnrv5kcsyxqa--azd-1744537379']
2025-04-13T09:57:37.2696622Z stderr F   File "<frozen importlib._bootstrap>", line 690, in _load_unlocked
2025-04-13T09:57:37.2696643Z stderr F   File "<frozen importlib._bootstrap_external>", line 940, in exec_module
2025-04-13T09:57:37.2696675Z stderr F   File "<frozen importlib._bootstrap>", line 241, in _call_with_frames_removed
2025-04-13T09:57:37.2696696Z stderr F   File "/usr/local/lib/python3.11/site-packages/opentelemetry/instrumentation/psycopg2/__init__.py", line 145, in <module>
2025-04-13T09:57:37.2696718Z stderr F     import psycopg2
2025-04-13T09:57:37.2696739Z stderr F ModuleNotFoundError: No module named 'psycopg2'
2025-04-13T09:57:37.2696760Z stderr F 2025-04-13 09:57:35,837 - INFO - Started server process [13]
2025-04-13T09:57:37.2696781Z stderr F 2025-04-13 09:57:35,838 - INFO - Waiting for application startup.

[pamelafox]

They say its fixed in the latest ?

open-telemetry/opentelemetry-python-contrib#3415 (comment)

Have you tried azure-monitor-opentelemetry==1.6.7 ?

[peinadoginesd]

I'm experiencing the same issue. Do you have any new in this regard?

[pamelafox]

1. They did actually just release a new version that fixes the psycopg2
2. However, I fixed app.py to only instrument for the libraries that we use in this app (no psycopg2) to avoid the issue entirely.

See:
https://github.com/Azure-Samples/azure-search-openai-demo/blob/main/app/backend/app.py#L812
And PR:
#2663

I will close this, please file a new issue if you see anything that I missed in that change.