You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Bug: get_parameters_by_name with raise_on_error=False still raises on transform errors in the decrypt path, and caches None #8594
With raise_on_error=False, a parameter whose transform fails is reported in _errors and does not stop the call, no matter whether it was fetched with GetParameter (decrypt) or GetParameters (batch). Failed transforms are not cached.
Current Behaviour
Two related problems in ssm.py:
_get_parameters_by_name_with_decrypt_option (:534) only catches GetParameterError, and BaseProvider.get hard-codes raise_on_transform_error=True, so a transform failure on a decrypted parameter raises TransformParameterError even in graceful mode. The same payload through the batch path returns None.
The batch path (:664) caches the None from a failed transform for max_age and does not list the name in _errors, while BaseProvider.get explicitly avoids caching None for that reason. A corrected parameter keeps returning None until the cache expires.
Catch TransformParameterError in the decrypt path and add the name to errors; in the batch path, skip add_to_cache when the transformed value is None and add the name to _errors.
Steps to Reproduce
Run the snippet above (it uses botocore.stub.Stubber, no AWS access needed).
Powertools for AWS Lambda (Python) version
latest
AWS Lambda function runtime
3.13
Packaging format used
PyPi
Debugging logs
Warning
This issue is ready to be picked up, by humans or by agents. If you want it, comment here first so we can assign it to you, and please take only one or two issues at a time so others can join in too. The reproduction above runs locally without AWS access. Thanks!
Hi! I'd like to work on this issue as my first open-source contribution.
I plan to investigate the decrypt-path transformation error handling and the caching of failed transformations, then propose a focused fix with regression tests for both cases.
Could you please confirm that this scope is appropriate and assign the issue to me before I start? Thank you!
Activity
Suraj1347 commentedon Oct 18, 2026
Hi! I'd like to work on this issue as my first open-source contribution.
I plan to investigate the decrypt-path transformation error handling and the caching of failed transformations, then propose a focused fix with regression tests for both cases.
Could you please confirm that this scope is appropriate and assign the issue to me before I start? Thank you!
leandrodamascena commentedon Oct 18, 2026
Hey @Suraj1347 pls go ahead.