Flask apps only signal an exception on real server errors #1611
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
connexion/apps/flask_app.py
Outdated
| @@ -85,6 +88,9 @@ def common_error_handler(self, exception): | |||
| headers=exception.get_headers(), | |||
| ) | |||
|
|
|||
| if is_server_error: | |||
There was a problem hiding this comment.
Can we replace this with if response["status"] >= 500?
There was a problem hiding this comment.
There is .status_code on the "ConnexionResponse" type.
Pull Request Test Coverage Report for Build 3469952874
💛 - Coveralls |
RobbeSneyders
approved these changes
Dec 7, 2022
RobbeSneyders
pushed a commit
that referenced
this pull request
Jan 24, 2023
Fixes the problem of non-exceptional "exceptions" being recorded by telemetry systems as a serious error. This expands on the changes made in #1326. The intention of that other change seems to be making telemetry systems like Sentry record serious errors. Diving into the Sentry implementation even shows that the signalled exception is recorded at "level" = "error". The root of the problem is that exceptions are being used for control flow, which is not ideal - convenient for app writers but not always for library maintainers. The connexion BadRequestProblem and NotFoundProblem, for example, should not be recorded as an error in a telemetry system. In my case [elastic-apm-python](https://github.com/elastic/apm-agent-python) is receiving these signals and recording 4xx events as serious errors. This pull request only propagates an exception signal to flask if it's a serious error. Aiohttp applications have a similar problem with exceptions being used for control flow - the problems middleware will convert the exception into an appropriate problem response but things like the elastic apm python telemetry middleware will see that exception and record it as a serious error. Interestingly aiohttp also uses exceptions for control flow and the elastic apm agent was patched to specifically ignore aio web exceptions below the 5xx status response range. Elastic apm and sentry cannot be expected to be aware of non-serious control flow exceptions used by various libraries though. So, a solution for Aiohttp applications is a separate problem. Co-authored-by: Enerqi <>
jcristau
added a commit
to jcristau/balrog
that referenced
this pull request
Jun 12, 2023
The connexion update includes spec-first/connexion#1611, so this hopefully Fixes: mozilla-releng#2870, mozilla-releng#2809
jcristau
added a commit
to jcristau/balrog
that referenced
this pull request
Jun 12, 2023
The connexion update includes spec-first/connexion#1611, so this hopefully Fixes: mozilla-releng#2870, mozilla-releng#2809
jcristau
added a commit
to jcristau/balrog
that referenced
this pull request
Jun 12, 2023
The connexion update includes spec-first/connexion#1611, so this hopefully Fixes: mozilla-releng#2870, mozilla-releng#2809
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes the problem of non-exceptional "exceptions" being recorded by telemetry systems as a serious error. This expands on the changes made in #1326. The intention of that other change seems to be making telemetry systems like Sentry record serious errors. Diving into the Sentry implementation even shows that the signalled exception is recorded at "level" = "error".
The root of the problem is that exceptions are being used for control flow, which is not ideal - convenient for app writers but not always for library maintainers. The connexion BadRequestProblem and NotFoundProblem, for example, should not be recorded as an error in a telemetry system. In my case elastic-apm-python is receiving these signals and recording 4xx events as serious errors.
This pull request only propagates an exception signal to flask if it's a serious error.
Aiohttp applications have a similar problem with exceptions being used for control flow - the problems middleware will convert the exception into an appropriate problem response but things like the elastic apm python telemetry middleware will see that exception and record it as a serious error. Interestingly aiohttp also uses exceptions for control flow and the elastic apm agent was patched to specifically ignore aio web exceptions below the 5xx status response range. Elastic apm and sentry cannot be expected to be aware of non-serious control flow exceptions used by various libraries though. So, a solution for Aiohttp applications is a separate problem.