Prep: Site for 1.2 release #2877

singhpk234 · 2025-10-23T16:46:37Z

Preps the website for the 1.2 release notes and artifact yet to be published

Update (10/24): The artifacts have been released as of yesterday

dimas-b

Re: DO NOT MERGE - mark the PR as "draft"?

site/hugo.yaml

dimas-b · 2025-10-23T17:31:24Z

site/content/downloads/_index.md

+    The realm-level feature flag `ALLOW_SETTING_SUB_CATALOG_RBAC_FOR_FEDERATED_CATALOGS` (default: true) controls whether this functionality can be enabled or modified at the catalog level.
+  - Added support for S3-compatible storage that does not have STS (use `stsUavailable: true` in catalog storage configuration)
+  - Added a Management API endpoint to reset principal credentials, controlled by the `ENABLE_CREDENTIAL_RESET` (default: true) feature flag.
+  - Added AWS CloudWatch Event Sink Implementation


Do we want to flag this as "beta"? If so, I think we can do that without re-releasing 1.2.0.

I do not recall if we discussed this before, so apologies for rocking the boat. Cf. https://lists.apache.org/thread/yx7pkgczl6k7bt4k4yzqrrq9gn7gqk2p

we can update the release notes in the website if we have concencus ? (post this pr as well, wdyt ?) we did that for 1.1 release

why do we wanna call it as beta though ? is it the interface gonna change or something ?

Hi @dimas-b, could you elaborate the reason here? I'm not sure I understanding the reason behind beta label and re-releasing?

The reason for "beta" is in the ML thread (linked above)... TL;DR: possible changes to binary / JSON representation.

If we're ok flagging persisted events as "beta" we could do this in the docs / release notes on the site, I think.

I also support flagging events as beta.

For a few reasons:

The API is still evolving, see this ML thread: https://lists.apache.org/thread/rl5cpcft16sn5n00mfkmx9ldn3gsqtfy

As the ML thread linked above shows, the SQL schema of the events table could evolve and require SQL migration.

Existing listeners are both incomplete:

JDBC: only processes after-create-table, after-create-catalog events, other events are ignored.

CloudWatch: only processes after-refresh-table events, other events are ignored.

As I already suggested a few times, a complete, general-purpose, production-ready event listener must either process all events equally, or expose a configuration option to select which events to process or ignore. See #2630 and #2836 (comment) for more discussions on this specific topic.

For all the above reasons, I think it would be nicer for our users to mark the events feature as beta/experimental, so that they don't accidentally rely on it prematurely in production-critical deployments.

Another point for "beta": the current Polaris Evolution doc has stronger guarantees for persisted data than for java interfaces.

If events only had a java representation, it would be not necessary to flag them as "beta" now. However, given that events are persisted and their representation in the Polaris database is very likely to change per dev ML discussions, I think we ought to warn uses about foreseeable upgrade issues in this area with the help of the "beta" label (plus some brief explanation of the issue).

CC: @jbonofre : related to you comment in the main thread.

dev ML discussion: https://lists.apache.org/thread/xogjyozb7qqj5jjs330f5s6lw69ldptv

singhpk234 · 2025-10-23T17:46:31Z

heads up vote passed in IPMC :)

site/content/downloads/_index.md

dimas-b · 2025-10-23T19:04:54Z

FYI: Hugo build is broken (on main) 😢

hugo_publish_apache-1  | ERROR [en] REF_NOT_FOUND: Ref "keycloak-idp.md": "/polaris/site/content/in-dev/unreleased/managing-security/external-idp/_index.md:254:95": page reference "keycloak-idp.md" is ambiguous
hugo_publish_apache-1  | ERROR [en] REF_NOT_FOUND: Ref "keycloak-idp.md": "/polaris/site/content/releases/1.2.0/managing-security/external-idp/_index.md:254:95": page reference "keycloak-idp.md" is ambiguous

singhpk234 · 2025-10-23T19:40:40Z

FYI: Hugo build is broken (on main) 😢

is it due to merging 1.2 in versioned-docs branch ? i just copied the in-dev/unreleased to release/1.2.0, hmm, any pointers ?

dimas-b · 2025-10-23T21:21:02Z

I'll start a dev ML thread... looks like a general CI issue 🤔

dimas-b · 2025-10-23T21:34:55Z

@singhpk234 : actually, I guess #2879 broke this... Would you mind rolling it back for now and we can redo it with more caution later? -- #2881

flyrain · 2025-10-23T21:47:15Z

hugo_publish_apache-1  | ERROR [en] REF_NOT_FOUND: Ref "keycloak-idp.md": "/polaris/site/content/in-dev/unreleased/managing-security/external-idp/_index.md:254:95": page reference "keycloak-idp.md" is ambiguous
hugo_publish_apache-1  | ERROR [en] REF_NOT_FOUND: Ref "keycloak-idp.md": "/polaris/site/content/releases/1.2.0/managing-security/external-idp/_index.md:254:95": page reference "keycloak-idp.md" is ambiguous
hugo_publish_apache-1  | ERROR [en] REF_NOT_FOUND: Ref "creating-a-catalog": "/polaris/site/content/in-dev/unreleased/getting-started/using-polaris/_index.md:34:35": page reference "creating-a-catalog" is ambiguous
hugo_publish_apache-1  | ERROR [en] REF_NOT_FOUND: Ref "creating-a-catalog": "/polaris/site/content/releases/1.2.0/getting-started/using-polaris/_index.md:34:35": page reference "creating-a-catalog" is ambiguous
hugo_publish_apache-1  | ERROR [en] REF_NOT_FOUND: Ref "telemetry-tools": "/polaris/site/content/in-dev/unreleased/telemetry.md:196:44": page reference "telemetry-tools" is ambiguous
hugo_publish_apache-1  | ERROR [en] REF_NOT_FOUND: Ref "telemetry-tools": "/polaris/site/content/releases/1.2.0/telemetry.md:196:44": page reference "telemetry-tools" is ambiguous

Yeah, we need to use relative paths for them

jbonofre

I'm not convinced we should absolutely flagged any feature as beta. I'm more in favor of informing the community it's available, and fix/improve in next releases if needed.
If it works, it works as it is today (that's the purpose of open source projects).

adutra · 2025-10-24T16:49:39Z

I'm not convinced we should absolutely flagged any feature as beta. I'm more in favor of informing the community it's available, and fix/improve in next releases if needed. If it works, it works as it is today (that's the purpose of open source projects).

But then we need to deal with angry users, we've seen this already: #2671.

I think being open and honest about the state of Polaris features is key in building a trust relationship with the community. For the events feature in particular, I think it's highly likely it will change substantially in the future.

singhpk234 · 2025-10-25T03:37:17Z

But then we need to deal with angry users, we've seen this already: #2671.

#2671 was caused due to a bug IMHO, which made user upgrade from 1.0.1 to 1.1.0 fail even without schema version upgrade. we do need to provide some automation for schema version upgrades !

I think being open and honest about the state of Polaris features is key in building a trust relationship with the community.

Agree with you on this ! but where are we being dishonest ? we are just saying hey this is something we have, try it out if you want to, if we put out a breaking change we should call it out in our release docs, reconsider that there is users using it and make change accordingly and this is true for any features we ship.

If the rationale is there is going to be breaking change for sure regarding events and the community agrees to (which my understanding is what your and @dimas-b feedback is). I propose i remove the CloudWatch sink from the release notes entirely (this will make this feature like NoSQL where parts of code is checked in but not complete / ready) and if we want we can mark CloudWatch sink or the whole EventListner in general as beta in the follow-up when we have concencus on thread @dimas-b started.
https://lists.apache.org/thread/xogjyozb7qqj5jjs330f5s6lw69ldptv

We already have release artifacts published in maven https://mvnrepository.com/artifact/org.apache.polaris/polaris-core as of (10/23), It would be great to move forward with this PR. Please let me know your thoughts ?

adutra · 2025-10-26T13:17:14Z

we are just saying hey this is something we have, try it out if you want to, if we put out a breaking change we should call it out in our release docs, reconsider that there is users using it and make change accordingly and this is true for any features we ship.

The above is not dishonest per se, but it's omitting two important messages:

The API is still evolving, including the SQL schema.
The implementation is currently incomplete.

I propose i remove the CloudWatch sink from the release notes entirely [...] if we want we can mark CloudWatch sink or the whole EventListner in general as beta in the follow-up.

AFAIK, both sinks (CloudWatch and SQL) are being introduced in 1.2.

Regardless of any mention in the changelog, this means that if we don't mark the API experimental now, they will be implicitly released as production-ready. It would be strange to downgrade this API from production-ready to beta in a future release imho.

dimas-b · 2025-10-27T15:56:48Z

I think we should try to make a complete message in the 1.2.0 docs / release notes upfront, reflecting the current status of features (specifically related to events). Updating these notices after they get published should be avoided from my POV, especially since we know about potential issues already.

Let's reach consensus on ML, and then revision the RNs: https://lists.apache.org/thread/xogjyozb7qqj5jjs330f5s6lw69ldptv

flyrain

Thanks @singhpk234!

singhpk234 · 2025-10-27T18:56:07Z

As recommended flagged Events added additional notes on events getting dropped. I have used word preview instead of beta, we have been using the word Experimental till now in 1.0.0 and EventListener is already marked as Experimental as of 1.0.0, please let me know what you think.

please take another pass @dimas-b @adutra @jbonofre when you get some time.

dimas-b · 2025-10-27T20:03:30Z

site/content/downloads/_index.md

+  - **Events Persistence (Preview)**: Introduced new event types and added support for persisting events to both Relational JDBC Persistence and AWS CloudWatch.
+
+    **Note**: This is a preview feature. The persistence schema is subject to change in future releases, and previously stored event data MAY become unreadable (i.e., dropped) after an upgrade.
+  - Generate Request IDs (if not specified) for all incoming requests to Polaris.


Automatic ID generation is removed by #2914 (assuming it gets merged).

I believe we should mention this as a "preview" too, and make a note that is may be removed in the next release.

I would not mind removing the Generate Request IDs point from the list of features completely. Per email discussion linked from #2914 it was not really intended to be a user-level feature.

thats make sense, i think the general concecus as per linked is to use Quarkus OTEL ! I removed the auto request id generation

dimas-b · 2025-10-27T20:04:40Z

The text about events LGTM 👍

update site

de857f0

github-project-automation bot added this to Basic Kanban Board Oct 23, 2025

github-project-automation bot moved this to PRs In Progress in Basic Kanban Board Oct 23, 2025

singhpk234 requested review from dimas-b and flyrain October 23, 2025 16:46

dimas-b reviewed Oct 23, 2025

View reviewed changes

singhpk234 marked this pull request as draft October 23, 2025 17:46

singhpk234 marked this pull request as ready for review October 23, 2025 17:50

address review feedback

7557580

sfc-gh-ygu reviewed Oct 23, 2025

View reviewed changes

site/content/downloads/_index.md Show resolved Hide resolved

Add extra line

3cc1d34

jbonofre previously approved these changes Oct 24, 2025

View reviewed changes

github-project-automation bot moved this from PRs In Progress to Ready to merge in Basic Kanban Board Oct 24, 2025

sfc-gh-prsingh dismissed jbonofre’s stale review via b51bcc1 October 27, 2025 18:49

mark event persistence as preview

bbc7f81

sfc-gh-prsingh force-pushed the feature/1.2-release branch from b51bcc1 to bbc7f81 Compare October 27, 2025 18:50

singhpk234 requested review from adutra, dimas-b and jbonofre October 27, 2025 18:50

flyrain previously approved these changes Oct 27, 2025

View reviewed changes

dimas-b reviewed Oct 27, 2025

View reviewed changes

remove request-id

8dd8fbb

sfc-gh-prsingh dismissed flyrain’s stale review via 8dd8fbb October 27, 2025 20:26

singhpk234 requested review from dimas-b and flyrain October 27, 2025 20:28

flyrain approved these changes Oct 27, 2025

View reviewed changes

singhpk234 merged commit 4e42502 into apache:main Oct 27, 2025
15 checks passed

github-project-automation bot moved this from Ready to merge to Done in Basic Kanban Board Oct 27, 2025

Prep: Site for 1.2 release #2877

Prep: Site for 1.2 release #2877

Uh oh!

Conversation

singhpk234 commented Oct 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dimas-b left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

dimas-b Oct 24, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

singhpk234 commented Oct 23, 2025

Uh oh!

Uh oh!

dimas-b commented Oct 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

singhpk234 commented Oct 23, 2025

Uh oh!

dimas-b commented Oct 23, 2025

Uh oh!

dimas-b commented Oct 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

flyrain commented Oct 23, 2025

Uh oh!

jbonofre left a comment

Choose a reason for hiding this comment

Uh oh!

adutra commented Oct 24, 2025

Uh oh!

singhpk234 commented Oct 25, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

adutra commented Oct 26, 2025

Uh oh!

dimas-b commented Oct 27, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

flyrain left a comment

Choose a reason for hiding this comment

Uh oh!

singhpk234 commented Oct 27, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

dimas-b commented Oct 27, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

singhpk234 commented Oct 23, 2025 •

edited

Loading

dimas-b Oct 24, 2025 •

edited

Loading

dimas-b commented Oct 23, 2025 •

edited

Loading

dimas-b commented Oct 23, 2025 •

edited

Loading

singhpk234 commented Oct 25, 2025 •

edited

Loading

dimas-b commented Oct 27, 2025 •

edited

Loading

singhpk234 commented Oct 27, 2025 •

edited

Loading

dimas-b commented Oct 27, 2025 •

edited

Loading