Propagate FunctionInvocationId for single-instance and bulk purge requests

Ouverte
  • Quelque chose ne va pas ?
#630 0 commentaires 0 réactions 0 personnes assignées Voir sur GitHub

@Fire162 y travaille déjà.

Depuis le 3/11/2026.

  • #631 par @Fire162 — ouverte

Évaluation

Difficulté
3/5
Temps estimé
1-2 jours
Accessibilité débutants
72/100
Type d'issue
Bug
Clarté
Clairement spécifiée
Activité
Active
Stack technique
python
Domaine
api

Piste de recherche

Start in azure/durable_functions/models/DurableOrchestrationClient.py at purge_instance_history and purge_instance_history_by, then read azure/durable_functions/models/utils/http_utils.py to understand the existing DELETE helper and header contract. Add regression coverage for both purge calls, including clients with distinct invocation IDs and without one; done means the outgoing header is correct and other purge behavior remains unchanged.

Rédigé par le modèle d'indexation à partir du texte de l'issue.

Description

Needs: Triage :mag:

Summary

Python Durable client purge operations do not forward the originating Azure Functions invocation ID, so the host cannot emit the correlated ClientOperationReceived logs needed by the remaining Python purge E2E assertions in Azure/azure-functions-durable-extension#3327.

This is a source-inspection report, not a deployed-application incident. The scope is the existing single-instance and bulk purge APIs, not adding new orchestration capabilities.

SDK line and next-generation tracking

This issue covers the stable Python v1 HTTP implementation in this repository; the current stable PyPI release is azure-functions-durable 1.7.0. The newer Python v2 Functions SDK is maintained in microsoft/durabletask-python under azure-functions-durable and has a published 2.0.0rc3 preview. Its separate gRPC invocation-metadata gap is tracked in microsoft/durabletask-python#314.

The extension's BasicPython app currently has an unversioned azure-functions-durable requirement, which normally selects the stable release rather than opting into v2 previews. Keep the two transport/version scopes distinct; maintainers can resolve the stable-line dependency through a backport or an explicit migration to a fixed newer SDK, rather than treating this issue as the only current Python implementation.

Current behavior and source evidence

Verified on dev at badf8016eb8186958ebd07908576deb1bb33e945:

Reproduction and expected behavior

Construct a DurableOrchestrationClient with a known, valid invocation ID, capture or mock its outgoing HTTP requests, and invoke each purge API with valid arguments. The current call sites omit the ID even though the DELETE helper supports it. Both outgoing requests should carry X-Azure-Functions-InvocationId equal to the originating invocation ID; purge URLs, filters, response parsing, and error behavior should remain unchanged.

Acceptance Criteria

  • Both purge APIs pass the client's invocation ID to the existing DELETE helper.
  • Regression tests cover single-instance and filtered bulk purge and verify the actual outgoing header or the helper argument plus the existing transport-header contract.
  • Tests cover distinct invocation IDs and clients without an invocation ID; no ID is fabricated or leaked between clients.
  • Existing purge behavior remains unchanged apart from the correlation header.
  • Record the published SDK version containing the fix in this issue so the extension E2E dependency can be updated and its Python PurgeInstances exemption removed.

Related

This issue tracks the upstream SDK dependency only. Enabling mandatory extension E2E assertions follows after the fixed package is available.

Langage dominant
Python
Étoiles
157
Forks
71
Aucune PR mergée en 30 j

Préparer son environnement

Ouvrir dans Codespaces

Lance le conteneur de développement du projet dans votre navigateur, avec votre propre compte GitHub.

Par où commencer

  1. Lisez l'issue en entier, puis le guide de contribution du projet.
  2. Signalez en commentaire que vous la prenez — cela évite que deux personnes fassent le même travail.
  3. Forkez le dépôt et travaillez sur une branche.
  4. Ouvrez une pull request qui référence le numéro de l'issue.

Autres issues de Azure/azure-functions-durable-python

Toutes les issues de Azure/azure-functions-durable-python

Issues similaires

Plus d'issues Python

Recevez les nouvelles issues par e-mail

Un résumé court des issues GitHub adaptées aux débutants.