Repository navigation
Status of testing Providers that were prepared on August 06, 2026 #71246
Description
I have a kind request for all the contributors to the latest provider distributions release.
Could you please help us to test the RC versions of the providers?
The guidelines on how to test providers can be found in
Verify providers by contributors
Let us know in the comments, whether the issue is addressed.
Important
This is an expedited release. The 2026-08-01 wave was cancelled because it
carried the pandas < 3 cap from #70791, since reverted in #71100 and replaced
by pandas 3 support in #71103. Every provider is re-cut, so the voting window is
shorter than usual — please test promptly if you can.
Ticks already carried over from the cancelled wave's issue (#70953) mark code
that has not changed; please do not re-verify those. Only the unticked items
need attention.
keycloak is deliberately not in this wave — #70550 depends on middleware
that is still being fixed (#71077, #70981). It will follow in a later release.
These are providers that require testing as there were some substantial changes introduced:
Provider airbyte: 6.0.1rc2
- Fix AirbyteJobSensor marking a cancelled job as success in deferrable mode (#69786): @steveahnahn
Linked issues:
Provider alibaba: 3.4.0rc2
- Add OSSRemoteLogIO.from_config and register oss remote logging scheme (#70682): @baha-bouali
Linked issues:
Provider amazon: 9.34.0rc2
- Cancel the Redshift statement when a user kills the deferred task (#69676): @steveahnahn
- Support setting an S3 object ACL for remote task logs (#70394): @FrankYang0529
Linked issues: - Add
DmsReloadTablesOperator(#70569): @AlejandroMorgante - Add deferrable mode to DynamoDBValueSensor (#69799): @harshhh817
Linked issues: - Refuse the team agnostic fall-through for a team scoped AWS secret name (#70878): @potiuk
- Retry EKS resource deletion on transient ResourceInUseException (#70260): @ramitkataria
- Apply SageMaker operator template-field checks after rendering (#70485): @mitre88
Linked issues: - Resolve S3ToRedshiftOperator connection and kwargs after template rendering (#70486): @mitre88
Linked issues: - Move AwsToAwsBaseOperator fallback out of init (#70445): @fat-catTW
Linked issues: - Fix EcsRunTaskOperator deferred logs read from the wrong region (#70474): @ColtenOuO
Linked issues: - Allow templated fields in AppFlow operators (#70440): @fat-catTW
Linked issues: - Fix EcsRunTaskOperator deferred-logs region and template field init (#70464): @fat-catTW
Linked issues: - Fix MongoToS3Operator aggregate-pipeline detection before rendering (#70330): @MannXo
Linked issues: - Strip GlueDataQualityOperator ruleset after rendering (#70331): @1fanwang
Linked issues: - Move template-field validation/transformation out of init for Amazon Bedrock operators (#70306): @Avcu
Linked issues: - Move template-field validation out of init for Amazon DataSync operator (#70324): @Avcu
Linked issues: - Only refuse team scoped like secret ids when multi_team is on (#71078): @amoghrajesh
- Keep S3DeleteObjectsOperator validation in init, narrow to ValueError (#70359): @AmoghAtreya
Linked issues: - Rename StepFunctionStartExecutionOperator's input field to match its constructor argument (#70544): @bramhanandlingala
Linked issues: - Fix mypy arg-type error in Redshift Data get_primary_keys (#70677): @Lee-W
- Add IMPORT_ERRORS_ALL permission for import errors of files with no registered Dag (#69790): @potiuk
Linked issues: - Move amazon dms operators template-field validation to execute() (#70450): @ahilashsasidharan
Linked issues: - Revert DMS provision checks to init with is not None polarity (#70634): @ahilashsasidharan
Linked issues: - Document effect of state-store cleanup for operators with durable execution (#70721): @amoghrajesh
- Remove hard-coded deferrable crawler run from example_glue system test (#70206): @seanghaeli
Linked issues:
Provider anthropic: 0.2.1rc2
- Move template-field validation out of AnthropicAgentSessionOperator init (#70432): @bramhanandlingala
Linked issues:
Provider apache.hive: 9.6.1rc2
Provider apache.kafka: 1.15.1rc2
- Fix Kafka consumer not being closed on error in ConsumeFromTopicOperator (#69641): @TimurRakhmatullin86
Linked issues: - Fix AwaitMessageTrigger crash on tombstone messages when apply_function is unset (#69665): @TimurRakhmatullin86
Linked issues: - Fix KafkaBaseHook.test_connection missing oauth_cb for managed Kafka (#69507): @TimurRakhmatullin86
- Fix AwaitMessageTriggerFunctionSensor not honoring timeout (#69639): @TimurRakhmatullin86
Linked issues: - Validate ProduceToTopicOperator topic after rendering (#70333): @1fanwang
Linked issues:
Provider apache.livy: 4.6.0rc2
Provider apache.spark: 6.3.1rc2
- Anchor spark-submit log trail on actual exception instead of lines (#70513): @amoghrajesh
- Spark: Fix missing space in YARN waitAppCompletion error message (#71055): @onlyarnav
Linked issues: - Decouple SparkSubmitOperator resumable deployment backends (#68679): @onlyarnav
Linked issues: - Drop unused post_submit_commands copy from SparkSubmitOperator init (#70323): @1fanwang
Linked issues:
Provider arangodb: 2.9.6rc2
Provider celery: 3.23.1rc2
- Honor verbose logging for Celery workers (#69828): @jx2lee
Linked issues:
Provider cncf.kubernetes: 10.21.0rc2
- Add
--min-completed-minutestocleanup-podsto prevent KPO race condition (#70595): @noamst-monday
Linked issues: - Add opt-in concurrent pod creation to KubernetesExecutor (#68480): @1fanwang
- Prevent deferrable
KubernetesPodOperatorlog parsing from blocking the triggerer event loop (#69661): @aeroyorch - Fix KubernetesPodOperator dry_run requiring live Kubernetes API client (#70234): @Aaryan123456679
Linked issues: - Build KubernetesInstallKueueOperator manifest URL after rendering (#70335): @1fanwang
Linked issues: - Mark KubernetesPodOperator and AgentOperator as durable capable (#70289): @amoghrajesh
- Validate KubernetesResourceBaseOperator yaml_conf after rendering (#70337): @1fanwang
Linked issues: - Document Kubernetes RBAC permissions (#68716): @dhkim1920
Linked issues: - Documenting effect if task state store cleanup for KPO (#70722): @amoghrajesh
Provider cohere: 1.6.7rc2
- Fix CohereEmbeddingOperator input_text normalization before rendering (#70325): @MannXo
Linked issues:
Provider common.ai: 0.7.0rc3
- Add pydantic-ai capability matrix references doc to Common AI (#69887): @gopidesupavan
- Support excluding files from
common.aiAgent Skills discovery (#69924): @kaxil - Use task state store for
common.aidurable execution on Airflow 3.3+ (#68926): @kaxil - Support .txt files in LLMFileAnalysisOperator (#70431): @ColtenOuO
- Support cloud URI globs in DocumentLoaderOperator (#70299): @guan404ming
- Support require_approval in LLMBranchOperator (#70651): @guan404ming
- Render multi-branch review choices as a multi-select (#71046): @guan404ming
- Allow Common AI SQL imports without DataFusion (#69990): @aaron-y-chen
- Fix HITL review showing altered agent output to reviewers (#70070): @guan404ming
- Fix LLMSQLQueryOperator not stripping single-line markdown code fences (#70137): @ColtenOuO
- Harden common.ai SQLToolset allowed_tables against function/COPY bypass (#70134): @kaxil
- Preserve output_type through human approval in LLM operators (#70075): @guan404ming
- Fix DocumentLoaderOperator ignoring unknown parser on byte input (#70071): @guan404ming
- Reject unsupported require_approval in common.ai operators (#70069): @guan404ming
- Fix
@task.llm_branchimport failure on Task SDK-only workers (#70068): @guan404ming - Fix
common.aidurable execution skipping Toolset-capability tools (#69881): @kaxil - Redact secrets from LLMRetryPolicy classification prompts (#70229): @Lee-W
- Validate common AI tool-call arguments to enable pydantic-ai retries (#70096): @zozo123
- Validate template fields after rendering in common.ai operators (#70338): @1fanwang
Linked issues: - Add dataclasses-json floor to common.ai llamaindex extra (#69755): @FrankYang0529
- Add a Retry Policies category to the provider registry (#70499): @Lee-W
- Remove noqa:S101 from production code (#70378): @Lee-W
- Mark KubernetesPodOperator and AgentOperator as durable capable (#70289): @amoghrajesh
- Add self-hosted model guide for the common.ai provider (#69867): @Lee-W
- Update retired model ids and fill doc gaps in common.ai provider docs (#69711): @Lee-W
- Add quick start guide to common.ai provider docs (#69552): @Lee-W
- Document the dynamic
system_promptpattern for common-ai agents (#69636): @jason810496 - Add an Examples entry point to the common.ai provider docs (#69650): @Lee-W
- Document the sql extra required for LLMSQLQueryOperator (#70564): @jedcunningham
Provider common.compat: 1.18.0rc2
- Gate audit log rows not tied to a Dag on a dedicated AccessView (#70759): @potiuk
- Add IMPORT_ERRORS_ALL permission for import errors of files with no registered Dag (#69790): @potiuk
Linked issues:
Provider common.sql: 2.1.0rc2
- Add SQLBulkLoadOperator to Common SQL Provider (#69362): @SameerMesiah97
- Fix SQLColumnCheckOperator crash on non-numeric column bounds (#70895): @shahar1
Linked issues: - Fix SQL check tolerance for negative expected values (#69736): @steveahnahn
Linked issues: - Fix SQL value checks for negative expected values (#69893): @aaron-y-chen
Provider databricks: 7.18.1rc2
- Review and update constraint dependencies (#70652): @potiuk
- Validate DatabricksSQLStatementsSensor statement fields after rendering (#70340): @1fanwang
Linked issues: - Validate DatabricksCopyIntoOperator template fields after rendering (#70339): @1fanwang
Linked issues: - Document effect of state-store cleanup for operators with durable execution (#70721): @amoghrajesh
- Docs: clarify Databricks retry_delay uses exponential backoff (#68622): @sduvvuri26
Provider dbt.cloud: 4.9.3rc2
- Fix DbtCloudRunJobOperator having false failures during deferred polling (#70581): @amoghrajesh
Linked issues: - Derive DbtCloudGetJobRunArtifactOperator default file name after rendering (#70336): @1fanwang
Linked issues:
Provider docker: 4.5.9rc2
Provider edge3: 4.3.0rc2
- Add queue parameter to edge list-workers subcommand (#70436): @shubhamraj-git
- edge3: Exit non-zero when worker admin CLI commands fail (#70112): @dkranchii
- Add team_name tag to remaining multi-team metrics (#69943): @Tegh25
Linked issues:- Linked Issue #68294: @stephen-bracken
Provider elasticsearch: 6.9.0rc2
- Add ElasticsearchRemoteLogIO.from_config and register elasticsearch scheme (#70525): @yuseok89
Linked issues: - Do not show @timestamp in logs in UI (#70790): @k-popov
- Clarify Elasticsearch remote-logging scheme needs Airflow 3.3.0+ (#70920): @jason810496
Linked issues:- Linked Issue #70525: @yuseok89
Provider fab: 3.8.0rc2
- Gate audit log rows not tied to a Dag on a dedicated AccessView (#70759): @potiuk
- Add IMPORT_ERRORS_ALL permission for import errors of files with no registered Dag (#69790): @potiuk
Linked issues: - Add
get_display_nametoBaseUserinterface (#70583): @pierrejeambrun
Linked issues:- Linked Issue #68833: @VladaZakharova
- Fix asset state store writes forbidden under FAB auth manager (#70478): @justinpakzad
- Stop dumping every Dag to stdout when syncing FAB permissions (#70939): @Eason09053360
Linked issues:- Linked Issue #53918: @uranusjr
Provider git: 0.4.2rc2
- git dag bundle to use the private key in submodule pull (#62938): @Pandry
- Use ValueError for conflicting Git connection key arguments (#70538): @mitre88
Linked issues:
Provider google: 22.3.0rc3
- Emit per-statement OpenLineage events for BigQuery script jobs (#69234): @aaron-y-chen
Linked issues:- Linked Issue #47579: @luke-hoffman1
- Add Vertex AI Agent Engine operators (Create, Get, Query, Update, Delete) (#68479): @AlejandroMorgante
- Adding durable execution to
BigQueryInsertJobOperator(#69542): @amoghrajesh - Register Stackdriver remote task logging handler (#70549): @fat-catTW
Linked issues: - Add CloudSQLNoOperationInProgressSensor for parallel admin ops (#68151): @Vamsi-klu
Linked issues:- Linked Issue #68040: @radhwene
- Linked Issue #68361: @radhwene
- Enable provider-based GCS remote logging resolution (#70504): @aaron-y-chen
Linked issues: - Make schema_fields templated in GCSToBigQueryOperator (#69108): @nikolauspschuetz
Linked issues:- Linked Issue #31481: @quantumlicht
- Fix Google Links for Sovereign Cloud (#69805): @molcay
- Fix Google Cloud Batch error messages to include job name (#69783): @parkhojeong
- Validate Gemini batch results_folder in the worker that writes results (#70393): @shahar1
Linked issues:- Linked Issue #70362: @daesung9511
- Fail deferred Cloud Composer tasks when the GCP operation errors (#70430): @SEPURI-SAI-KRISHNA
- Apply the team scope in the Google Secret Manager backend (#70869): @potiuk
- Delay BigQueryToMsSqlOperator source table parsing (#70493): @fat-catTW
Linked issues: - Emit GCSToGCSOperator deprecation warnings after rendering (#70449): @haseebmalik18
Linked issues: - Emit GCSListObjectsOperator delimiter deprecation warning after rendering (#70533): @yuseok89
Linked issues: - Move Compute Engine validation out of constructors (#70454): @daesung9511
Linked issues: - Prevent premature validation in Gemini batch job operators (#70362): @daesung9511
Linked issues: - Only refuse team scoped like secret ids when multi_team is on (#71078): @amoghrajesh
- Strip only the literal properties/ prefix from GA property link IDs (#70919): @Eason09053360
- Update system test to the latest api version of google-ads (#69865): @olegkachur-e
- Drop redundant BigQueryInsertJobOperator job ID snapshot in init (#70494): @fat-catTW
Linked issues: - Document effect of state-store cleanup for operators with durable execution (#70721): @amoghrajesh
Provider hashicorp: 4.8.0rc2
Provider imap: 3.12.1rc2
- reject mixed-separator path traversal in imap attachment names (#70665): @Samin061
- check imap attachment symlink on the resolved destination path (#69194): @Samin061
Provider microsoft.azure: 14.1.0rc2
- Add WasbRemoteLogIO.from_config and register wasb remote logging scheme (#70301): @Andrushika
Linked issues: - Apply the ambiguous-id refusal to get_config in the Key Vault backend (#70899): @potiuk
Linked issues: - Refuse the team agnostic fall-through for a team scoped Key Vault secret name (#70876): @potiuk
- Fix duplicated task logs in the WASB log handler (#70860): @bmanan7
Linked issues:- Linked Issue #67144: @jvstein
- Linked Issue #70867: @bmanan7
- Linked Issue #68304: @jvstein
- Validate AzureVirtualMachineStateSensor target_state after rendering (#70329): @1fanwang
Linked issues: - Fail AzureVirtualMachineStateSensor fast on an invalid target_state (#70372): @shahar1
Linked issues: - Only refuse team scoped like secret ids when multi_team is on (#71078): @amoghrajesh
- Use ValueError for unsupported recursive glob in ADLS upload (#70540): @mitre88
Linked issues:
Provider neo4j: 3.12.1rc2
Provider openai: 1.8.2rc2
- Use monotonic clock for OpenAIBatchTrigger polling timeout (#69534): @YAshhh29
Linked issues:- Linked Issue #69408: @YAshhh29
- Validate trigger events in Openai deferrable tasks (#69506): @takayoshi-makabe
Linked issues:- Linked Issue #41554: @josix
- Linked Issue #69379: @takayoshi-makabe
Provider openlineage: 2.20.0rc3
- OpenLineage: add execute_in_thread to emit task events without forking (#68708): @mobuchowski
Linked issues:- Linked Issue #67115: @AutomationDev85
- Linked Issue #66572: @potiuk
- Linked Issue #66573: @potiuk
- Linked Issue #66574: @potiuk
- Linked Issue #65714: @mobuchowski
- Add team_name tag to OpenLineage metrics (#69109): @SameerMesiah97
Linked issues:- Linked Issue #68996: @ferruzzi
- Fix openlineage team name handling, add dag_version_data and TI note (#70772): @kacpermuda
Linked issues:- Linked Issue #70760: @kacpermuda
- Linked Issue #70452: @SameerMesiah97
- OpenLineage: reuse per-process adapter in dag-state-change pool workers (#69283): @gang-zh
Linked issues:- Linked Issue #69284: @gang-zh
- Fix OpenLineage team_name access and other failing paths (#70626): @kacpermuda
- Bump openlineage libraries to 1.52 in openlineage provider (#70773): @kacpermuda
- Document GlueJobOperator support for OpenLineage Spark injection (#69652): @mobuchowski
Provider opensearch: 1.12.0rc1
- Add OpensearchRemoteLogIO.from_config and register opensearch scheme (#70295): @FrankYang0529
Linked issues: - Do not show @timestamp in logs in UI (#70790): @k-popov
Provider papermill: 3.13.2rc2
Provider postgres: 7.0.1rc2
- Use built-in exceptions for Postgres hook input validation (#70537): @mitre88
Linked issues: - Document psycopg3 driver switch for PostgresHook connections (#70360): @shahar1
Linked issues:- Linked Issue #69089: @Dev-iL
- Linked Issue #69526: @Dev-iL
Provider sftp: 6.0.1rc2
- Fix encrypted private keys in async SFTP hook (#70858): @fallintoplace
- Support Paramiko 5 in SSH and SFTP providers (#69712): @shivaam
Linked issues:- Linked Issue #69669: @shivaam
Provider smtp: 3.0.3rc2
Provider snowflake: 6.16.0rc2
- Add SnowflakeCortexAgentOperator to Snowflake Provider (#69939): @SameerMesiah97
- Log check_query_output response via format string in Snowflake SQL API hook (#67848): @michaelneely
- Fix snowflake provider installs picking an unimportable pyOpenSSL (#71145): @rjgoyln
- Move SnowparkContainerJobOperator template-field validation to execute (#70332): @justinpakzad
Linked issues: - Rewrite SnowparkContainerJobOperator exclusivity check and move back … (#70874): @ahilashsasidharan
Linked issues: - Document effect of state-store cleanup for operators with durable execution (#70721): @amoghrajesh
Provider ssh: 6.0.1rc2
- Fix SSHRemoteJobOperator cleanup failing with custom remote_base_dir (#69885): @harshhh817
Linked issues:- Linked Issue #69813: @AchimGaedkeLynker
- quote remote job paths in posix ssh command builders (#70091): @Samin061
- Keep SSH remote job running when its PTY session hangs up (#70573): @jason810496
- Fix template-field validation timing in ssh provider operators (#70315): @bramhanandlingala
Linked issues: - Support Paramiko 5 in SSH and SFTP providers (#69712): @shivaam
Linked issues:- Linked Issue #69669: @shivaam
Provider standard: 1.17.0rc2
- WaitSensor - time_to_wait templated (#70480): @raphaelauv
- Read HITLOperator subject/body into the summary lazily (#70345): @1fanwang
Linked issues: - Make the HITLOperator summary extension point public (#70391): @shahar1
Linked issues: - Give each task its own start trigger arguments in standard sensors (#70888): @potiuk
- Deprecating
poll_intervalinExternalTaskSensorforpoke_interval(#68997): @jroachgolf84
Linked issues:- Linked Issue #68990: @surendarvarmas
- Fix BashOperator script detection after templating (#70369): @fat-catTW
Linked issues: - Fix DateTimeSensor crash when target_time renders to a datetime (#70320): @1fanwang
Linked issues: - Validate TriggerDagRunOperator logical_date after template rendering (#70490): @mitre88
Linked issues:
Provider teradata: 3.6.2rc2
- Fix BteqOperator declaring template_fields as a bare string (#70862): @Eason09053360
- Fix Teradata transfer template field initialization (#70392): @fat-catTW
Linked issues:
Provider weaviate: 3.4.1rc2
- Fix WeaviateIngestOperator input_data validation before rendering (#70326): @MannXo
Linked issues: - Fix duplicated and incorrect words in documentation (#70868): @Sdnsoumy
Provider yandex: 4.5.1rc2
- Refuse the team agnostic fall-through for a team scoped Lockbox secret name (#70877): @potiuk
- Only refuse team scoped like secret ids when multi_team is on (#71078): @amoghrajesh
All users involved in the PRs:
@shivaam @rjgoyln @guan404ming @gang-zh @Avcu @aeroyorch @bmanan7 @TimurRakhmatullin86 @shubhamraj-git @jroachgolf84 @yuseok89 @baha-bouali @ahilashsasidharan @Pandry @bramhanandlingala @kaxil @amoghrajesh @Andrushika @kacpermuda @harshhh817 @1fanwang @fallintoplace @Sdnsoumy @molcay @raphaelauv @noamst-monday @seanghaeli @fat-catTW @Samin061 @AlejandroMorgante @gopidesupavan @jason810496 @k-popov @SEPURI-SAI-KRISHNA @FrankYang0529 @Vamsi-klu @sduvvuri26 @mobuchowski @dkranchii @zozo123 @mitre88 @ColtenOuO @potiuk @AmoghAtreya @Lee-W @jx2lee @Eason09053360 @nikolauspschuetz @haseebmalik18 @daesung9511 @onlyarnav @olegkachur-e @takayoshi-makabe @steveahnahn @YAshhh29 @MannXo @shahar1 @dhkim1920 @ramitkataria @Aaryan123456679 @aaron-y-chen @jedcunningham @SameerMesiah97 @pierrejeambrun @Tegh25 @BasPH @justinpakzad @parkhojeong @michaelneely
Activity
potiuk commentedon Aug 6, 2026
The issue body above mentions more contributors than GitHub renders as links — only the first 50 become real mentions, so the rest of you were never notified. Tagging you here so you see it.
This is an expedited release: the 2026-08-01 wave was cancelled and every provider re-cut, so the voting window is shorter than usual. If you have a change in this wave, please test promptly.
Items already ticked were carried over from #70953 and cover unchanged code — no need to re-verify those.
@Tegh25 @stephen-bracken @yuseok89 @timestamp @k-popov @pierrejeambrun @VladaZakharova @Eason09053360 @uranusjr @Pandry @luke-hoffman1 @Vamsi-klu @radhwene @nikolauspschuetz @quantumlicht @molcay @parkhojeong @daesung9511 @SEPURI-SAI-KRISHNA @haseebmalik18 @olegkachur-e @BasPH @Samin061 @Andrushika @bmanan7 @jvstein @YAshhh29 @takayoshi-makabe @josix @mobuchowski @AutomationDev85 @ferruzzi @kacpermuda @gang-zh @Dev-iL @fallintoplace @shivaam @michaelneely @rjgoyln @AchimGaedkeLynker @raphaelauv @jroachgolf84 @surendarvarmas @Sdnsoumy
Drafted-by: Claude Code (Opus 5); reviewed by @potiuk before posting
onlyarnav commentedon Aug 6, 2026
#71053 was an output rendering change
conflictswith->conflicts withVerified it, no testing required
SEPURI-SAI-KRISHNA commentedon Aug 6, 2026
Tested
google 22.3.0rc3for #70430 (Fail deferred Cloud Composer tasks when the GCP operation errors) — works as expected.Installed the RC from PyPI into a clean venv (
apache-airflow-providers-google==22.3.0rc3, resolved againstapache-airflow3.3.0) and verified against the installed artifact rather than a source checkout.Unit tests — the trigger tests ship green against the RC:
Behavioural check with real LRO objects — the unit tests use a stand-in, so I also exercised the trigger against actual
google.longrunning.operations_pb2.Operationvalues:done=Truewitherror.messagesetAirflowException: Cloud Composer Environment error: quota exceededdone=True, no error{'operation_name': ..., 'operation_done': True}done=Falsetwice, then failsFor contrast, the pre-fix loop on that same failed
Operationreturns{'operation_name': 'op', 'operation_done': True}— a success event for a failed operation, which is the bug. The RC no longer does this.I don't have a live Composer environment, so this is verification against the released artifact rather than a real environment creation/update failure.
Drafted-by: Claude Code (Opus 5); reviewed by @SEPURI-SAI-KRISHNA before posting
fallintoplace commentedon Aug 6, 2026
Tested
apache-airflow-providers-sftp==6.0.1rc2for #70858.Installed the provider from PyPI in a clean Python 3.13 virtualenv with Airflow 3.3.0. The environment resolved Paramiko 5.0.0 and AsyncSSH 2.24.0.
The documented
private_key_passphraseextra was passed correctly toasyncssh.import_private_key. I also tested an encrypted private key against a temporary local AsyncSSH server and successfully opened an SFTP session.The existing
passphraseextra behavior still works, and the shared Paramiko 5 RSA private-key validation also passed.No issues found.
takayoshi-makabe commentedon Aug 7, 2026
Tested apache-airflow-providers-openai==1.8.2rc2 for #69506.
Installed Airflow 3.3.0 + the RC into a clean venv from PyPI and exercised OpenAITriggerBatchOperator.execute_complete / validate_execute_complete_event / OpenAIBatchTrigger.run() directly (no live OpenAI API needed, this is pure trigger-event branching logic).
As a control, I also installed the pre-fix 1.8.1 and confirmed all three bugs actually reproduce there:
status="cancelled"batch_idOpenAIBatchJobExceptionevent=NoneTypeError: 'NoneType' object is not subscriptableOpenAITriggerEventErrorstatus(e.g."expired")OpenAITriggerEventErrorstatus="success"/"error"Also confirmed OpenAIBatchTrigger.run() now yields exactly one terminal event (no spurious duplicate error event after a terminal event).
All checks pass on the RC. No issues found.
Drafted-by: Claude Code; reviewed by @takayoshi-makabe before posting
seanghaeli commentedon Aug 7, 2026
Verified #70206
k-popov commentedon Aug 7, 2026
Verified #70790
radhwene commentedon Aug 7, 2026
Tested apache-airflow-providers-google==22.3.0rc3 for #68151 (CloudSQLNoOperationInProgressSensor).
I installed the RC from PyPI in a clean Python 3.12 environment with Airflow 3.3.0. pip check is clean, including the RC dependencies on common-sql and common-compat.
I also compared it with Google provider 19.5.0 to confirm that the new sensor, trigger, status constant and CloudSQLHook.list_operations() are all present only in the RC.
For poke(), I checked the main paths:
I also confirmed that the existing Cloud SQL import and export operators still import correctly.
I did not run a real parallel workload against Cloud SQL, so this validates the released artifact and its synchronous sensor behaviour, not the remaining timing window under live concurrency.
No issues found in the scope tested.
⸻
Drafted with Claude ; reviewed by @radhwene before posting.
michaelneely commentedon Aug 7, 2026
Verified #67848
The result of a SnowflakeSqlApiOperator is now correctly displayed:
YAshhh29 commentedon Aug 7, 2026
Tested
apache-airflow-providers-openai==1.8.2rc2for #69534.I installed the PyPI wheel in a clean Python 3.10.20 environment with
Airflow 3.1.3, OpenAI provider 1.8.2rc2, and OpenAI SDK 2.53.0.
pip checkis clean. The wheel SHA-256 isc7b9e56d38b3edd914780e11a94e048c7905dfaf70999b7c2c1365d89c63f570,matching the PyPI artifact.
For #69534, I verified that:
OpenAITriggerBatchOperatorpassestimeouttoOpenAIBatchTrigger.timeoutand legacyend_timeforms both serialize correctly.100 -> 101 -> 106, the trigger reportstimeout after six elapsed seconds.
the same simulated jump.
I also confirmed that the dependency change from #69408 remains present:
the RC metadata requires plain
openai>=2.37.0without thedatalibextra. The clean environment contains no
numpy,pandas, orpandas-stubs. This metadata change was already present in provider 1.8.1,so this part is regression verification rather than an RC-versus-1.8.1
difference.
RC artifact checks: 9 passed.
Pre-fix comparison: 1 passed.
I used the clean-virtualenv contributor verification method. No live
OpenAI API call was made; this validates the published artifact's metadata,
serialization, validation, operator handoff, and timeout behavior.
No issues found in the tested scope.
Drafted-by: GitHub Copilot; reviewed by @YAshhh29 before posting
Eason09053360 commentedon Aug 7, 2026
Verified my three items on this wave, all against the RC wheels from PyPI
installed into the official
apache/airflow:3.1.3-python3.10imagefab 3.8.0rc2 — Stop dumping every Dag to stdout when syncing FAB permissions #70939: with 82 example Dags serialized in the DB,
airflow sync-perm --include-dagsnow prints only the two expectedteradata 3.6.2rc2 — Fix BteqOperator declaring template_fields as a bare string #70862:
BteqOperator.template_fields == ("sql",),andsql="SELECT '{{ ds }}';"actually renders toSELECT '2026-08-07';google 22.3.0rc3 — Strip only the literal properties/ prefix from GA property link IDs #70919: a property named
properties/s123now persistsproperty_id="s123". The previouslstrip("properties/")ate the leadingsand persisted123.All three look good to me :)
Drafted-by: Claude Code (Opus 5); reviewed by @Eason09053360 before posting
molcay commentedon Aug 7, 2026
Thank you very much @potiuk for the RC.
From google provider perspective, we did NOT find any regression.
We are giving +1.
potiuk commentedon Aug 7, 2026
google 22.3.0rc3 is removed from the 2026-08-06 wave
The google provider has been pulled from this wave following a binding -1 on the
vote thread:
apache-airflow-providers-google==22.3.0rc3cannot be installed onPython 3.14.
Root cause is upstream, not in the provider code.
google-cloud-aiplatformcapsits
evaluationextra atlitellm>=1.83.7,<1.86.0, and no litellm in that rangeinstalls on Python 3.14, so
pip install apache-airflow-providers-googlefailswith
ResolutionImpossibledespite the provider advertisingrequires_python >= 3.10. It affects the already-released 22.2.x line too.google-cloud-aiplatformwithout theevaluationextra, declares that extra's dependencies directly, and splits thelitellm requirement by Python version.
What this means for this issue
and testing/verification on them continues to count.
not be promoted when this wave is published.
against code that is unchanged by Workaround bug in google-claude-aiplatform evaluation extra (Python 3… #71273, which only touches dependency
metadata. Please do not re-verify them.
What happens to google
google will be re-cut as 22.3.0rc4 in a separate google-only wave with release
preparation date 2026-08-08, carrying #71273 on top of the same code. That wave
gets its own testing status issue, and the ticks from here will be carried over to
it.
Drafted-by: Claude Code (Opus 5); reviewed by @potiuk before posting
dhkim1920 commentedon Aug 8, 2026
Verified #68716
apache-airflow-providers-cncf-kubernetes==10.21.0rc2with Airflow 3.3.0 on Kind Kubernetes v1.30.13.I re-ran the RBAC scenarios covered by #68716 using restricted ServiceAccounts, including Pod and Job operators,
deferrable Jobs, KubernetesExecutor with multi-namespace mode, cleanup, SparkApplication, generic custom resources,
and Kueue installation/Jobs.
All expected operations succeeded, while out-of-scope permissions were denied. I found no rc2-specific issues, and
#40210 appears addressed from the documented RBAC perspective.
Drafted-by: Codex (GPT-5); reviewed by @dhkim1920 before posting
fat-catTW commentedon Aug 8, 2026
Verified the following Amazon provider changes with
apache-airflow-providers-amazon==9.34.0rc2on Airflow 3.3.0 in clean venvs. I also compared each relevant behavior withapache-airflow-providers-amazon==9.33.0.#70445:
AwsToAwsBaseOperatordestination connection fallbackChecked
dest_aws_conn_idfallback behavior whendest_aws_conn_idis omitted:9.33.0: aftersource_aws_conn_idchanges from'source_before_render'to'source_after_render',dest_aws_conn_idstill returns'source_before_render'9.34.0rc2:dest_aws_conn_idfollows the updated source and returns'source_after_render'Also checked that explicit
dest_aws_conn_id=Noneand templateddest_aws_conn_idare preserved during operator construction.#70440: templated validation fields in AppFlow operators
Checked
AppflowRunDailyOperatorwith templatedsource,source_field, andfilter_date.Previous release
9.33.0rejects the operator during construction:Source {{ params.source }} is not supported for AppflowRunDailyOperator!RC
9.34.0rc2allows construction, renders the fields, validates after rendering, and executes the patched run path with:source='salesforce'source_field='CreatedDate'filter_date='2022-05-26T00:00:00+00:00#70464: deferred ECS logs region selection
Checked
EcsRunTaskOperator.execute_complete()whenregion_name='task-region'andawslogs_region='logs-region'.Previous release
9.33.0callsAwsLogsHookwithregion_name='task-region'.RC
9.34.0rc2callsAwsLogsHookwithregion_name='logs-region'and returns the expected log output.Result: all three work as expected.
Drafted-by: Codex (GPT-5); reviewed by @fat-catTW before posting
potiuk commentedon Aug 9, 2026
The 2026-08-06 wave has been released. 44 providers are now available on PyPI, in SVN and their documentation is live at https://airflow.apache.org/docs/
Thank you all for testing and verifying the release candidates!
Two providers were excluded from this wave and re-cut in the 2026-08-08 wave:
google22.3.0 — did not install on Python 3.14 (Restore google-cloud-aiplatform[evaluation] on Python 3.14 once the litellm ceiling is lifted #71268, upstream Litellm limits are too constraining for Python 3.14 googleapis/python-aiplatform#7057, worked around in Workaround bug in google-claude-aiplatform evaluation extra (Python 3… #71273). Now22.3.0rc4.cncf.kubernetes10.21.0 —version_addedfor the concurrent pod creation options pointed at 10.20.0 instead of 10.21.0 (Fix version_added for the concurrent pod creation options #71157). Now10.21.0rc3.Testing for those two continues in #71321.
Drafted-by: Claude Code (Opus 5) (no human review before posting)
shivaam commentedon Aug 9, 2026
Tested four changes that were not already reported in the issue comments, using exact PyPI RC artifacts in clean venvs with Airflow 3.3.0; all passed:
apache-airflow-providers-sftp==6.0.1rc2andapache-airflow-providers-ssh==6.0.1rc2naturally resolved Paramiko 5.0.0, imported from the installed artifacts, and validated RSA private keys withrsa-sha2-512. The 6.0.0 controls retain the Paramiko<5cap and reproduce the oldKeyError: 'ssh-rsa'when Paramiko 5 is deliberately forced.2.9.6rc2raises the intended built-inValueErrorfor a missing collection operation; the 2.9.5 control raisesAirflowException, and valid insert dispatch remains intact.0.4.2rc2raises the intended built-inValueErrorfor conflicting connection keys; the 0.4.1 control raisesAirflowException, and single-key construction remains intact.3.12.1rc2includes deprecatedsqlin the standard template/rendering pipeline, renders it, and passes the rendered query to the hook; 3.12.0 was used as the control.I also ran a real Airflow 3.3.0 Breeze deployment with the exact RC artifacts installed from PyPI and mounted provider sources disabled. A focused DAG parsed without import errors and completed successfully through the scheduler and LocalExecutor: the artifact-identity gate and the behavior tasks for the four items above passed. The API health check reported the metadata DB, scheduler, triggerer, and DAG processor healthy. No live external provider services or credentials were used.
No issues found.
Drafted-by: OpenAI Codex (GPT-5.6); reviewed by @shivaam before posting
Vamsi-klu commentedon Aug 22, 2026