davidzollo opened a new issue, #11356:
URL: https://github.com/apache/seatunnel/issues/11356

   ## Background
   Issue #11239 and PR #11301 add an important first step for upgrade 
compatibility: a cross-version restore workflow with representative scenarios.
   
   That work is worth landing, but it should be treated as **phase 1**, not as 
the full upgrade-compatibility story.
   
   SeaTunnel still needs a clearer release-grade answer to this question:
   
   **What upgrade guarantees do we actually verify before shipping changes that 
affect persisted runtime state?**
   
   ## Problem to solve
   An initial restore workflow is useful, but it does not yet cover the broader 
compatibility problem.
   
   Remaining gaps include:
   - scenario coverage is still intentionally small
   - not all high-risk state surfaces are represented
   - not every risky code area has a clear policy for when upgrade checks 
should run
   - failure output still needs to be actionable for maintainers
   - the community still lacks a written compatibility boundary for what is 
guaranteed and what is best effort
   
   ## Proposed scope
   Track phase-2 work that turns upgrade compatibility from a single workflow 
into a maintainable compatibility program.
   
   Suggested scope:
   - define the minimum supported upgrade scenarios that must stay green on 
`dev`
   - expand scenario coverage for stateful source, sink, CDC, and 
checkpoint/restore paths where it materially reduces regression risk
   - define when the workflow runs: scheduled only, manually triggered, or 
selectively required for high-risk changes
   - make failures clearly identify whether the break happened during state 
creation, restore, resume, or post-restore validation
   - document the compatibility contract that the workflow is asserting
   
   ## Why this needs STIP first
   This is not just a CI scripting task. It affects:
   - release policy
   n- CI cost and runtime budget
   - which modules/changes are treated as compatibility-sensitive
   - what level of backward compatibility SeaTunnel is publicly willing to claim
   
   Because of that, **the claimant should submit a STIP design first and get 
maintainer agreement before starting implementation**.
   
   The STIP should explicitly define:
   - the supported version matrix and scenario-selection strategy
   - the minimum high-risk paths that must be covered
   - trigger/gating policy for `dev` and for PRs touching risky areas
   - artifact/version sourcing and reproducibility rules
   - how compatibility failures are triaged and who owns follow-up
   
   ## Acceptance criteria
   - SeaTunnel has a documented upgrade-compatibility policy instead of only an 
ad hoc workflow.
   - Phase-2 coverage extends beyond the initial scenarios from #11239 / PR 
#11301.
   - Failure output is actionable enough for maintainers to identify the broken 
stage quickly.
   - English and Chinese docs are updated where compatibility guarantees are 
user-visible.
   
   ## Non-goals for the first version
   - exhaustive coverage of every connector and every historical version
   - turning every PR into a full compatibility matrix run immediately
   - claiming stronger guarantees than the workflow actually verifies
   
   ## Related work
   - Issue #11239
   - PR #11301
   - Issue #11353
   - Issue #11354
   


-- 
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]

Reply via email to