amoghrajesh opened a new pull request, #67833:
URL: https://github.com/apache/airflow/pull/67833
<!-- SPDX-License-Identifier: Apache-2.0
https://www.apache.org/licenses/LICENSE-2.0 -->
<!--
Thank you for contributing!
Please provide above a brief description of the changes made in this pull
request.
Write a good git commit message following this guide:
http://chris.beams.io/posts/git-commit/
Please make sure that your code changes are covered with tests.
And in case of new features or big changes remember to adjust the
documentation.
Feel free to ping (in general) for the review if you do not see reaction for
a few days
(72 Hours is the minimum reaction time you can expect from volunteers) - we
sometimes miss notifications.
In case of an existing issue, reference it using one of the following:
* closes: #ISSUE
* related: #ISSUE
-->
---
##### Was generative AI tooling used to co-author this PR?
<!--
If generative AI tooling has been used in the process of authoring this PR,
please
change below checkbox to `[X]` followed by the name of the tool, uncomment
the "Generated-by".
-->
- [x] Yes claude sonnet 4.6
<!--
Generated-by: [Tool Name] following [the
guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#gen-ai-assisted-contributions)
-->
### What problem are we solving?
AIP-103 introduced two new storage features — per-task-instance key/value
storage and
per-asset key/value storage. Both were named `task_state` and `asset_state`
respectively.
As we tested this more and more, the naming started getting ambigous:
Airflow already has a
well-established concept of "task state" meaning the lifecycle phase of a
task instance (`RUNNING`, `SUCCESS`, `FAILED`,etc.). Using the same word for a
separate storage feature caused confusion in code review,
documentation, and conversations and it was unclear whether `task_state` in
any given context
referred to "is this task running?" or "what did this task store?".
### Current behaviour
Before this change, the two features were entangled in naming:
- `context["task_state"]` — the dict key for accessing the key/value store
in `execute()`
- `context.ti.state` — the lifecycle state of the task instance
Both used the word "state". Reviewers regularly had to ask for
clarifications. Test names, URL paths
(`/execution/state/ti/`), CLI commands (`cleanup-task-states`), ORM model
classes
(`TaskStateModel`, `AssetStateModel`), and SDK comms messages
(`GetTaskState`, `SetTaskState`)
all perpetuated the ambiguity.
### Proposed change
Rename the key/value *storage* feature from `*_state` to `*_store`
throughout the codebase.
The naming follows the convention used in Kafka Streams, which calls this
pattern a
**State Store** — persistent storage attached to a processing unit,
explicitly distinguished
from the processing unit's lifecycle status.
| Layer | Before | After |
|---|---|---|
| **Context keys** | `context["task_state"]`, `context["asset_state"]` |
`context["task_store"]`, `context["asset_store"]` |
| **ORM models** | `TaskStateModel`, `AssetStateModel` | `TaskStoreModel`,
`AssetStoreModel` |
| **DB modules** | `models/task_state.py`, `models/asset_state.py` |
`models/task_store.py`, `models/asset_store.py` |
| **Execution API routes** | `/execution/state/ti/`, `/state/asset/` |
`/execution/store/ti/`, `/store/asset/` |
| **Core API routes** | `/public/state/task/`, `/state/asset/` |
`/public/store/task/`, `/store/asset/` |
| **SDK comms messages** | `GetTaskState`, `SetTaskState`, `ClearTaskState`,
… | `GetTaskStore`, `SetTaskStore`, `ClearTaskStore`, … |
| **SDK accessor classes** | `TaskStateAccessor`, `AssetStateAccessors` |
`TaskStoreAccessor`, `AssetStoreAccessors` |
| **SDK client operations** | `client.task_state.get/set/delete` |
`client.task_store.get/set/delete` |
| **CLI subcommand** | `airflow state-store cleanup-task-states` | `airflow
state-store cleanup-task-store` |
| **Example Dags** | `example_task_state`, `example_asset_state` |
`example_task_store`, `example_asset_store` |
| **`ResumableJobMixin`** | `context["task_state"]` |
`context["task_store"]` |
**Not renamed:** `TaskState`/`TaskInstanceState` (lifecycle enum),
`task_states` parameters in
sensors/triggers (allowed lifecycle states), `GetTaskStates` (queries
*other* TIs' lifecycle
states)
### User Interaction change
For any operator using AIP-103 task/asset store in `execute()`:
- `context["task_state"]` → `context["task_store"]`
- `context["asset_state"]` → `context["asset_store"]`
- `def execute(self, context, asset_state=None)` → `asset_store=None`
(injected parameter name)
- Execution API clients calling `/execution/state/ti/` or `/state/asset/`
must update to
`/execution/store/ti/` and `/store/asset/`
AIP-103 has not shipped in a stable release, so no migration shim is
provided.
---
* Read the **[Pull Request
Guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#pull-request-guidelines)**
for more information. Note: commit author/co-author name and email in commits
become permanently public when merged.
* For fundamental code changes, an Airflow Improvement Proposal
([AIP](https://cwiki.apache.org/confluence/display/AIRFLOW/Airflow+Improvement+Proposals))
is needed.
* When adding dependency, check compliance with the [ASF 3rd Party License
Policy](https://www.apache.org/legal/resolved.html#category-x).
* For significant user-facing changes create newsfragment:
`{pr_number}.significant.rst`, in
[airflow-core/newsfragments](https://github.com/apache/airflow/tree/main/airflow-core/newsfragments).
You can add this file in a follow-up commit after the PR is created so you
know the PR number.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
