)]}'
{
  "log": [
    {
      "commit": "6910b0c339b2b8697c0bec4bc392263cec450aec",
      "tree": "bf636247e750390ad404b0a92187f3a7f781611d",
      "parents": [
        "0c06f6c0b9452d3799bdf70f49badef9e77e7444"
      ],
      "author": {
        "name": "Chareonwit Kunna",
        "email": "21arenabreakout12@gmail.com",
        "time": "Sun Nov 22 14:30:41 2026 +0700"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 22 13:00:41 2026 +0530"
      },
      "message": "docs: fix doubled word in README (#6181)"
    },
    {
      "commit": "0c06f6c0b9452d3799bdf70f49badef9e77e7444",
      "tree": "8c43a35a3f964ce791883eb947598bfbcb181471",
      "parents": [
        "e04248704babd8e4848348152454471e291a48a5"
      ],
      "author": {
        "name": "Subham",
        "email": "subhamsangwan26@gmail.com",
        "time": "Sun Nov 22 03:35:07 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 21 16:05:07 2026 -0600"
      },
      "message": "fix(cli): summarize empty tables without snapshots metadata (#6161)"
    },
    {
      "commit": "e04248704babd8e4848348152454471e291a48a5",
      "tree": "f92fc31135e0ce21fb8aa03719d6dcaffa2bcfe2",
      "parents": [
        "5c38f0077b47767a47d102da56695bdd36647360"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Sat Nov 21 05:45:18 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 21:45:18 2026 -0800"
      },
      "message": "Update dependency com.diffplug.spotless:spotless-plugin-gradle to v8.12.0 (#6173)"
    },
    {
      "commit": "5c38f0077b47767a47d102da56695bdd36647360",
      "tree": "59974410873786c56cec2d00129106f669ebfaff",
      "parents": [
        "207abcdc1776326cadc50e2fd3bca7ca0c66213e"
      ],
      "author": {
        "name": "Adnan Hemani",
        "email": "adnan.h@berkeley.edu",
        "time": "Fri Nov 20 18:43:15 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 18:43:15 2026 -0800"
      },
      "message": "Split lineage privileges into table and view variants (#5574)\n\nPrivilege subsumption ignores the leaf entity type, so a single lineage\nprivilege implied by both TABLE_* and VIEW_* grants would let a view grant\non a namespace confer lineage access on every table in it. Add\nTABLE_READ_LINEAGE, TABLE_INGEST_LINEAGE, VIEW_READ_LINEAGE and\nVIEW_INGEST_LINEAGE, each implied only by grants on its own entity type,\nand split the operations to match, adding REFERENCE_LINEAGE_INPUT_VIEW.\n\nAdd the matching access types and implied grants to the shipped Ranger\nservice definition, with fixture cases covering them."
    },
    {
      "commit": "207abcdc1776326cadc50e2fd3bca7ca0c66213e",
      "tree": "e916a43e2b518994fafb0565f549525536773360",
      "parents": [
        "fcc1d476dd33e0a5f875c33a9cafbd2d21cb76c6"
      ],
      "author": {
        "name": "Danica Fine",
        "email": "danica.fine@gmail.com",
        "time": "Fri Nov 20 18:40:33 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 18:40:33 2026 -0800"
      },
      "message": "Fix CHANGELOG.md to correctly show changes in 1.9.0 and current main (#6158)\n\n* fix CHANGELOG.mg to show 1.9.0 changes\n\n* move mis-tagged 1.8.0 items to 1.9.0"
    },
    {
      "commit": "fcc1d476dd33e0a5f875c33a9cafbd2d21cb76c6",
      "tree": "670f176e4ec3b1db8514ee6fc59e41708c09fd5a",
      "parents": [
        "dc078b991c740d39c170579ec615c7fe6cd5e585"
      ],
      "author": {
        "name": "Alejandro Pajaron",
        "email": "apjorquera@gmail.com",
        "time": "Sat Nov 21 02:16:21 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 20:16:21 2026 -0500"
      },
      "message": "Add a guide for using the Polaris REST APIs with curl (#6157)\n\n* Add a guide for using the Polaris REST APIs with curl\n\nUsers often ask how to try out Polaris with plain HTTP requests, without\nthe Polaris CLI or a query engine. The docs had only a few isolated curl\nsnippets. None of them showed the whole flow.\n\nAdd a \"REST API with curl\" guide that walks through the Management and\nIceberg REST APIs step by step:\n- get an access token for the root principal\n- create a catalog that uses FILE storage, so no cloud account or\n  object store is needed\n- create a principal, a principal role and a catalog role, link them,\n  and grant CATALOG_MANAGE_CONTENT\n- as that principal, read the catalog config, then create, list and\n  load namespaces and tables\n- clean up in an order that works: a catalog can only be dropped once\n  its custom catalog roles are deleted, and purging a dropped table\n  needs DROP_WITH_PURGE_ENABLED\n\nThe guide comes with its own docker-compose file (Polaris only, in-memory\nmetastore). The larger request bodies are in\nsite/content/guides/assets/rest-api/ and are passed to curl with -d @file.\nThe guide-testing CI job runs the guide like any other, and hidden checks\nconfirm that the table is created and the cleanup completes.\n\nThe \"Connecting Using REST APIs\" section of the Using Polaris docs now\nlinks to the new guide.\n\nFixes #1119\n\n* Address review comments on the REST API guide\n\n- Extend the hidden cleanup check to verify that quickstart_user and\n  quickstart_user_role were deleted, so a failing DELETE no longer\n  goes unnoticed in CI.\n- Clarify that the config prefix is the catalog name, which is why the\n  following paths start with quickstart_catalog."
    },
    {
      "commit": "dc078b991c740d39c170579ec615c7fe6cd5e585",
      "tree": "ac1ee872fa148685245a9e684d140a882721d310",
      "parents": [
        "8dbc4db9af94e4e41eb89e7ffe9f062069d647dc"
      ],
      "author": {
        "name": "Alejandro Pajaron",
        "email": "apjorquera@gmail.com",
        "time": "Sat Nov 21 02:14:01 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 20:14:01 2026 -0500"
      },
      "message": "Support Iceberg procedures in the Polaris Spark catalog (#6136)\n\nThe Polaris SparkCatalog wraps Iceberg\u0027s SparkCatalog but did not\nimplement ProcedureCatalog, so any `CALL \u003ccatalog\u003e.system.\u003cprocedure\u003e`\nfailed because the catalog was not recognized as a procedure catalog.\n\nImplement ProcedureCatalog in the Spark 3.5 and 4.0 SparkCatalog and\ndelegate loadProcedure to the wrapped Iceberg catalog, the same way\nviews are already delegated. Spark 3.5 uses Iceberg\u0027s\norg.apache.spark.sql.connector.iceberg.catalog.ProcedureCatalog, and\nSpark 4.0 uses Spark\u0027s own org.apache.spark.sql.connector.catalog.ProcedureCatalog.\n\nProcedures are bound to the wrapped Iceberg catalog, so they apply to\nIceberg tables only, and table arguments should not include the\ncatalog name (e.g. \u0027db.tbl\u0027 rather than \u0027catalog.db.tbl\u0027).\n\nFixes #2550"
    },
    {
      "commit": "8dbc4db9af94e4e41eb89e7ffe9f062069d647dc",
      "tree": "22b715070df74560e6516a19c5d0e4089091d0fb",
      "parents": [
        "7ffdf1d4b85b929bd5607ce1211bca78ca17a654"
      ],
      "author": {
        "name": "Akash R Nilugal",
        "email": "akashnilugal@gmail.com",
        "time": "Sat Nov 21 06:42:20 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 20:12:20 2026 -0500"
      },
      "message": "Docs: fix CLI update and policies list examples (#6132)\n\nThe principals, principal-roles and catalog-roles  examples used\n--property, but those commands take --set-property. The\nexample was missing the required --namespace. Verified each old and new\ncommand against the CLI parser."
    },
    {
      "commit": "7ffdf1d4b85b929bd5607ce1211bca78ca17a654",
      "tree": "b9be8a7ea24185f624c8c87fe33b104c40930e0d",
      "parents": [
        "040e37feeee0cf0579cd6edb8685ec8a72db9031"
      ],
      "author": {
        "name": "youngrae kim",
        "email": "160102021+youngrae81@users.noreply.github.com",
        "time": "Sat Nov 21 10:10:55 2026 +0900"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 20:10:55 2026 -0500"
      },
      "message": "Evaluate the OPTIMIZED_SIBLING_CHECK readiness check with each realm\u0027s effective settings (#6129)\n\n* Evaluate the OPTIMIZED_SIBLING_CHECK readiness check with each realm\u0027s effective settings\n\nThe production readiness check for OPTIMIZED_SIBLING_CHECK read both flags from\na realm\u0027s override map only. A realm runs with the default for every key it\ndoes not override, so a realm that enabled the check and inherited\nALLOW_OPTIMIZED_SIBLING_CHECK from the defaults was refused at startup, while a\nrealm that inherited the enabled check but overrode the acknowledgement to\nfalse was not reported.\n\nThe check now resolves each flag per realm the way DefaultConfigurationStore\ndoes, override first and default otherwise, matching the GCS attribution check\nin the same class.\n\nFixes #6128\n\n* Skip realms that override neither flag and tighten the disable test\n\nA realm that overrides neither OPTIMIZED_SIBLING_CHECK nor\nALLOW_OPTIMIZED_SIBLING_CHECK runs with the defaults, which are already\nchecked, so it no longer gets a second error pointing to a property it does\nnot set.\n\nThe test for a realm disabling an inherited check now also withdraws the\nacknowledgement, so it only passes if the realm\u0027s OPTIMIZED_SIBLING_CHECK\noverride is honored."
    },
    {
      "commit": "040e37feeee0cf0579cd6edb8685ec8a72db9031",
      "tree": "666ffaa740c0401797e693a529d005740185394c",
      "parents": [
        "c63b8b55ae6c45ea4fda04ec8da65b90546e7499"
      ],
      "author": {
        "name": "Yufei Gu",
        "email": "yufei@apache.org",
        "time": "Fri Nov 20 11:46:17 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 11:46:17 2026 -0800"
      },
      "message": "Fix shell variable expansion in the GCS catalog example (#6153)"
    },
    {
      "commit": "c63b8b55ae6c45ea4fda04ec8da65b90546e7499",
      "tree": "9d750363d2b1015bb36babae9f379dc14d848e56",
      "parents": [
        "ef7960ad71ab8690b2c7b03dcc9e1ad6b0d035f6"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Fri Nov 20 16:14:13 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 21:44:13 2026 +0530"
      },
      "message": "Update github/codeql-action action to v4.39.4 (#6164)"
    },
    {
      "commit": "ef7960ad71ab8690b2c7b03dcc9e1ad6b0d035f6",
      "tree": "1c748d834c632b018238556f7c4f23ddc3fcfd43",
      "parents": [
        "80d88dbc9906ab3c3742e5f951f983bcb9ae9073"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Fri Nov 20 13:17:56 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 13:17:56 2026 +0000"
      },
      "message": "Update dependency org.mongodb:mongodb-driver-sync to v5.15.0 (#6156)"
    },
    {
      "commit": "80d88dbc9906ab3c3742e5f951f983bcb9ae9073",
      "tree": "8a9bd424c9c7794aa03d1fcb60b0dd85745cca93",
      "parents": [
        "76d651e985ff83cfe1d2e5682501e9d2be0ed045"
      ],
      "author": {
        "name": "Ayush Saxena",
        "email": "ayushsaxena@apache.org",
        "time": "Fri Nov 20 10:16:58 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 10:16:58 2026 +0530"
      },
      "message": "Fix commitTransaction leaking metadata files when a later table in the transaction fails (#6008)\n\ncommitTransaction writes each table\u0027s new metadata file inside doCommit, before\nthe single atomic entity update that publishes them all. In transaction mode the\nentity write is buffered rather than persisted, so doCommit\u0027s own cleanup\ndeliberately keeps the file: the transaction, not the individual commit, decides\nwhether it ever becomes reachable.\n\nThe transaction\u0027s cleanup only ran when that final entity update reported\nfailure. Any other failure escaping the loop over the tables, such as a\nrequirement that no longer holds, a rejected client-specified location or a\nlocation validation failure, propagated out of commitTransaction with the\nmetadata files already written for the earlier tables still on storage,\nreferenced by no entity. Nothing reclaims them, because\nwrite.metadata.delete-after-commit only walks the previous files of a metadata\nthat did get committed. The common trigger is a requirement that a concurrent\nwriter has invalidated, this endpoint\u0027s ordinary failure mode, which fails the\nrequest with 409.\n\nClean up on the same rule the single-table commit path applies in doCommit:\nevery failure before the persistence call has been attempted, and after that\nonly the failures the catalog considers safe, through\nshouldCleanupMetadataOnCommitFailure. Nothing is persisted while the per-table\nchanges are applied, so a failure there orphans the files already written\nwhatever its type, including the ones that predicate does not recognise, such as\nan update naming a schema that does not exist. Once the persistence call has\nbeen attempted the predicate decides, and withholds cleanup when the commit\noutcome is unknown, where deleting a file that a published pointer may reference\nis unrecoverable.\n\nThe cleanup that the failing final entity update performed now runs through that\nsame path: it is reported as CommitFailedException, which the predicate admits,\nso the behaviour on that path is unchanged."
    },
    {
      "commit": "76d651e985ff83cfe1d2e5682501e9d2be0ed045",
      "tree": "5abd8c0c7d6e9fea24bd05c8995123b16eca2bc1",
      "parents": [
        "a6ceec4f13019a95789520d9324db73982f184f1"
      ],
      "author": {
        "name": "Yufei Gu",
        "email": "yufei@apache.org",
        "time": "Thu Nov 19 20:05:37 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 20:05:37 2026 -0800"
      },
      "message": "Align the applicable policy response example with the REST API (#6142)"
    },
    {
      "commit": "a6ceec4f13019a95789520d9324db73982f184f1",
      "tree": "27f0a1257012d102e0585e7d2af9962c9f05fa2f",
      "parents": [
        "e45891b36abedae3a3d5726688f25db00df11105"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Fri Nov 20 04:31:22 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 20 04:31:22 2026 +0100"
      },
      "message": "NoSQL: Add Cassandra driver configuration and session support (#6088)\n\nA Cassandra persistence backend needs connection configuration that works with Polaris configuration sources and supports both standalone use and Quarkus deployments.\n\nAdd the Cassandra module and Java-driver dependencies, typed driver-option mapping, and programmatic session creation.\nIntegrate CDI-managed sessions and authentication providers, the Quarkus TLS Registry, and Micrometer metrics.\nValidate conflicting configuration and manage session and authentication-provider lifecycles.\n\nThis establishes the connection layer for the Cassandra backend.\n\nRelated to #6024, the illustrative Cassandra commit-stack PR."
    },
    {
      "commit": "e45891b36abedae3a3d5726688f25db00df11105",
      "tree": "33ed960a059203153a6f87295fa59d4af5a8cec4",
      "parents": [
        "779fce623a5ac3ae71d5e7ff409519e3e053a916"
      ],
      "author": {
        "name": "Prithvi S",
        "email": "prithvisivasankar@gmail.com",
        "time": "Fri Nov 20 05:48:15 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 19:18:15 2026 -0500"
      },
      "message": "Resolve catalog via getResolvedReferenceCatalogEntity in GET /v1/config handler (#6012)\n\nSigned-off-by: Prithvi S \u003cprithvisivasankar@gmail.com\u003e"
    },
    {
      "commit": "779fce623a5ac3ae71d5e7ff409519e3e053a916",
      "tree": "eda4ca587f7877db271150efed6e8fe088603144",
      "parents": [
        "c3e9f8dae26f3391f4e6e510ac26d24794470475"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Thu Nov 19 22:13:50 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 14:13:50 2026 -0800"
      },
      "message": "Update registry.access.redhat.com/ubi9/openjdk-21-runtime Docker tag to v1.24-5 (#6155)"
    },
    {
      "commit": "c3e9f8dae26f3391f4e6e510ac26d24794470475",
      "tree": "d36a7c1c6815f3747cf363cd884a1f509f2ec9df",
      "parents": [
        "cbe2037727ead43e5906c6636eb03b34af020aba"
      ],
      "author": {
        "name": "Yufei Gu",
        "email": "yufei@apache.org",
        "time": "Thu Nov 19 13:29:14 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 13:29:14 2026 -0800"
      },
      "message": "Clarify shell variable assignments in the Azure catalog example (#6154)"
    },
    {
      "commit": "cbe2037727ead43e5906c6636eb03b34af020aba",
      "tree": "0c4f0e9eefa45c7daf322b5c4d1b25338f7663da",
      "parents": [
        "b66027a86294e9a84522672ffb657258258f10be"
      ],
      "author": {
        "name": "Yufei Gu",
        "email": "yufei@apache.org",
        "time": "Thu Nov 19 10:27:36 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 10:27:36 2026 -0800"
      },
      "message": "Fix shell variable expansion in the catalog creation example (#6140)"
    },
    {
      "commit": "b66027a86294e9a84522672ffb657258258f10be",
      "tree": "123a1d6f026d0bf58a8ef18941e72ee231b4c058",
      "parents": [
        "f74ddfceea0e8fbbd9c926be9a420dc1a254273c"
      ],
      "author": {
        "name": "ssvynarchuk",
        "email": "286896404+ssvynarchuk@users.noreply.github.com",
        "time": "Thu Nov 19 20:27:13 2026 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 10:27:13 2026 -0800"
      },
      "message": "Remove obsolete Kafka startup retry (#6152)"
    },
    {
      "commit": "f74ddfceea0e8fbbd9c926be9a420dc1a254273c",
      "tree": "78b0d2db9f1873646c2faec5b04394700be860b5",
      "parents": [
        "2e953098816b78b91c5b3f1222c955b47f3157ef"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Thu Nov 19 18:12:49 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 23:42:49 2026 +0530"
      },
      "message": "Update dependency ch.qos.logback:logback-classic to v1.6.8 (#6149)"
    },
    {
      "commit": "2e953098816b78b91c5b3f1222c955b47f3157ef",
      "tree": "9f2b8c7d252ff3b14ac6bf1125e61020e972d83e",
      "parents": [
        "eee7e82f537868ddd1a15d683d2d382696bd52fb"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Thu Nov 19 17:50:38 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 23:20:38 2026 +0530"
      },
      "message": "Update dependency io.swagger:swagger-annotations to v1.6.17 (#6150)"
    },
    {
      "commit": "eee7e82f537868ddd1a15d683d2d382696bd52fb",
      "tree": "4acb4803780dbdb6f090fa4407a186763f9af55e",
      "parents": [
        "58245cce1bf3e5a3b6d99fac07608fa7522f31e0"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Thu Nov 19 17:10:49 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 22:40:49 2026 +0530"
      },
      "message": "Update dependency io.micrometer:micrometer-bom to v1.18.0 (#6151)"
    },
    {
      "commit": "58245cce1bf3e5a3b6d99fac07608fa7522f31e0",
      "tree": "04e4441e44345f9feb667f66dbb4ab0f8950a8d9",
      "parents": [
        "6362d2c405dc14fd7f4509646de5023cac6d7eed"
      ],
      "author": {
        "name": "Yufei Gu",
        "email": "yufei@apache.org",
        "time": "Thu Nov 19 09:02:35 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 09:02:35 2026 -0800"
      },
      "message": "Correct the PostgreSQL connection properties in the JDBC guide (#6139)"
    },
    {
      "commit": "6362d2c405dc14fd7f4509646de5023cac6d7eed",
      "tree": "f644066cfbed1a6d368711aa9d07299a8eb8bc5c",
      "parents": [
        "25e0281f20070c737d5882ad6f172aefc68d51db"
      ],
      "author": {
        "name": "Yuewei Zhou",
        "email": "91304878+ZephyrYWZhou@users.noreply.github.com",
        "time": "Thu Nov 19 08:31:06 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 08:31:06 2026 -0800"
      },
      "message": "Persistence: DynamoDB (atomic) listing and hasChildren (#6016)"
    },
    {
      "commit": "25e0281f20070c737d5882ad6f172aefc68d51db",
      "tree": "76179be937924d7ab51f1adac69d4bc46b4b83be",
      "parents": [
        "4b036243d5e698c0439936bdc92da66ec13ed103"
      ],
      "author": {
        "name": "Rishi",
        "email": "sririshindra@gmail.com",
        "time": "Thu Nov 19 07:37:59 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 10:37:59 2026 -0500"
      },
      "message": "Spec: Add shared pagination definitions for Polaris APIs (#6028)\n\n* Spec: Add shared pagination definitions for Polaris APIs\n\nNew Polaris list operations either reference the Iceberg REST Catalog\npagination definitions or define their own. The Iceberg definitions\nrequire a server to return all results when a request has no pageToken,\nwhich defeats the purpose of paginating, and per-operation copies drift\napart.\n\nAdd spec/polaris-pagination.yaml with one pageToken/pageSize parameter\npair and a page token schema, usable from both the catalog and the\nmanagement APIs. Its rules: any response may be paginated, neither\nparameter is needed to start, the server decides the page size, and a\nserver may return everything in one response. Each response names the\ntoken field in its own API\u0027s style.\n\nNo operation references the file yet. Existing operations keep their\ncurrent definitions.\n\n* Spec: Scope the pagination rules to operations that use them\n\nThe first rule now says that it applies to the responses of operations\nthat use these definitions, so it cannot be read as a statement about\nevery Polaris list operation. Operations that reference the Iceberg REST\nCatalog pagination definitions keep those rules.\n\n* Spec: Clarify page size, page token reuse and errors in pagination rules\n\npageSize is now an upper bound for a server that paginates, capped by the\nserver\u0027s own maximum, instead of a hint. That makes the separate \"the\nserver decides the page size\" rule redundant, so it is removed.\n\npageToken now says that an empty token means the first page, that a\nclient sends the same parameters on every page except pageSize, and that\nthe server returns 400 for a token it cannot use and may return it when\nthe other parameters change.\n\nThe header comment no longer lists the APIs or the response field names;\neach API spec names its own field.\n\n* Spec: Allow empty pages and treat an empty page token as the end\n\npageSize no longer refers to a server\u0027s \"maximum page size\", which can\nmean a number of entries or a number of bytes. It now says that a server\nmay return fewer results than pageSize, even none, in a page that is not\nthe last, and that a client follows the page token of an empty page.\n\nSince an empty pageToken in a request means no token, an empty page token\nin a response now also marks the end of the results, along with a missing\nor null one."
    },
    {
      "commit": "4b036243d5e698c0439936bdc92da66ec13ed103",
      "tree": "3fceae6d4ea33e48686da934fb96355adc44bc47",
      "parents": [
        "faec52afbbb9361bf3b547042b78344b48b52081"
      ],
      "author": {
        "name": "gracechen09",
        "email": "119986098+gracechen09@users.noreply.github.com",
        "time": "Thu Nov 19 10:28:38 2026 -0500"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 19 10:28:38 2026 -0500"
      },
      "message": "OPA: add operator-defined static properties to the authorization input context (#5925)\n\n* add opa context properties\n\n* add example usage in opa.md\n\n* simplify tests\n\n* add OPA context properties to application.properties\n\n* update example usage of opa context properties in opa.md\n\n* Pass context properties to new OpaPolarisAuthorizer call sites from main"
    },
    {
      "commit": "faec52afbbb9361bf3b547042b78344b48b52081",
      "tree": "1895b4fbeb9a87583634bf31064d9418308a1c49",
      "parents": [
        "0943b1b543ad6f257ffce77a2d6888f989ae816e"
      ],
      "author": {
        "name": "austent",
        "email": "11274313+deepdishgary@users.noreply.github.com",
        "time": "Wed Nov 18 16:35:05 2026 -0600"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 17:35:05 2026 -0500"
      },
      "message": "Add the R2_LOCAL_SIGNING credential vending mechanism for Cloudflare R2 (#5972)\n\n* Add the R2_LOCAL_SIGNING credential vending mechanism for Cloudflare R2\n\n* Carry the storage name in the R2 cache key and redact the parent secret\n\n* Report a lost R2 parent token as a server error\n\n* Note the shared parent token and SRCFG00050 in the changelog\n\n* Move the R2 changelog entries out of the 1.9.0 notes\n\n* Apply the new Java formatting to the R2 extension\n\n* Move the R2 changelog entries to a changelog.d fragment"
    },
    {
      "commit": "0943b1b543ad6f257ffce77a2d6888f989ae816e",
      "tree": "17cba84aca467411545fa02b4470c49212321003",
      "parents": [
        "7087773483bef9a3cb43c5748b1b9e0e62b8acc7"
      ],
      "author": {
        "name": "Akash R Nilugal",
        "email": "akashnilugal@gmail.com",
        "time": "Thu Nov 19 03:31:51 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 17:01:51 2026 -0500"
      },
      "message": "Return 400 instead of NPE when updating a catalog without currentEntityVersion (#6131)\n\nPUT /catalogs/{name} with a body that omits currentEntityVersion failed\nwith a NullPointerException (HTTP 500). PolarisAdminService.updateCatalog\ncompared the primitive int entity version to the nullable Integer from\nUpdateCatalogRequest, which unboxed null.\n\nUpdateCatalogRequest does not mark currentEntityVersion as required in the\nspec, unlike the Principal, PrincipalRole and CatalogRole update requests.\nReject the request with a BadRequestException instead, so the optimistic\nlock is never skipped."
    },
    {
      "commit": "7087773483bef9a3cb43c5748b1b9e0e62b8acc7",
      "tree": "17c483e55ecceae3a23105afbfb5cf571042cf82",
      "parents": [
        "888f98cce336248597289eba002baf08a7c3b530"
      ],
      "author": {
        "name": "colivetree",
        "email": "cefoliv@gmail.com",
        "time": "Wed Nov 18 21:43:13 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 13:43:13 2026 -0800"
      },
      "message": "fix(catalog): discard builder changes when filtering snapshots\u003drefs (#5811)"
    },
    {
      "commit": "888f98cce336248597289eba002baf08a7c3b530",
      "tree": "f25743032c8a29a05ad665d4d8bf4f7a74e15508",
      "parents": [
        "53a1b60125c4d61ec5d7a372ef9b18e008ee2e13"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Wed Nov 18 21:31:12 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 13:31:12 2026 -0800"
      },
      "message": "Update cockroachdb/cockroach Docker tag to v26.4.1 (#6135)"
    },
    {
      "commit": "53a1b60125c4d61ec5d7a372ef9b18e008ee2e13",
      "tree": "d39a63269be663cc1ea8ec6153bddc58b7d28e93",
      "parents": [
        "21f775e90392fa5da89bb1dec1e7694ffda18cf8"
      ],
      "author": {
        "name": "ssvynarchuk",
        "email": "286896404+ssvynarchuk@users.noreply.github.com",
        "time": "Wed Nov 18 23:14:06 2026 +0200"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 21:14:06 2026 +0000"
      },
      "message": "Support custom metadata paths for views (#6036)\n\n* Fix scoped credentials for custom view metadata paths\n\nInclude write.metadata.path when scoping view storage credentials and\nvalidate custom metadata paths against the same location restrictions\nused for tables. Record view locations for sibling overlap checks and\ninclude custom metadata paths during cross-namespace rename validation.\nApply structured restrictions to legacy views without a stored base\nlocation before their first update persists one.\n\nAdd create, update, removal, allowed-location, legacy-view, and rename\nregressions, including shared FILE and IAM-enforced S3 integration\ncoverage. Document custom metadata paths and view overlap checks.\n\n* Preserve legacy view layouts and scope custom metadata purge\n\nApply structured location restrictions to new views and path changes while\nallowing existing views to commit with unchanged layouts. Continue checking\nallowed locations and disabled storage types before writing metadata, and\navoid redundant persistence-time validation of legacy layouts.\n\nInclude custom view metadata paths in purge task credential scopes. Restore\ndefault-configuration coverage and add regressions for legacy updates,\nreplacement, storage-type restrictions, and scoped metadata deletion.\n\n* Validate view purge against recorded storage locations\n\n* Document view metadata path restrictions and cover sibling overlap\n\n* Simplify recorded view purge locations and clarify regression tests\n\n* Clarify legacy view compatibility and overlap consequences\n\n* Simplify view location collection and validation\n\n* Remove obsolete table-like update validation flag"
    },
    {
      "commit": "21f775e90392fa5da89bb1dec1e7694ffda18cf8",
      "tree": "cdf6ea641806ecdc4e8a3dcd9002376a18c817d8",
      "parents": [
        "51123ab818ced2327633d3fa36d2863ae0a84be7"
      ],
      "author": {
        "name": "Matt Faltyn",
        "email": "76570855+mattfaltyn@users.noreply.github.com",
        "time": "Wed Nov 18 11:03:14 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 14:03:14 2026 -0500"
      },
      "message": "Preserve idempotent retries across same-table registration (#5792)\n\n* Preserve idempotent retries across same-table registration\n\nFixes #5789\n\n* Extract helper for carrying forward idempotency keys\n\nShare the key-window copy between registerTable and table commits, and\nnote that re-registering older same-UUID metadata replays rolled-back\nupdate keys as success."
    },
    {
      "commit": "51123ab818ced2327633d3fa36d2863ae0a84be7",
      "tree": "a4ff7f42ec197c4f3935a951ce9162c8ca7feacc",
      "parents": [
        "306935471d34c7b1df8b291c043d4ca847ed55fd"
      ],
      "author": {
        "name": "JB Onofré",
        "email": "jbonofre@apache.org",
        "time": "Wed Nov 18 18:13:25 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 10:13:25 2026 -0800"
      },
      "message": "Add Polaris Community Meeting 2026-11-12 (#6127)"
    },
    {
      "commit": "306935471d34c7b1df8b291c043d4ca847ed55fd",
      "tree": "bda2f70eab8d1258fec6ff1999fd6376fed0910f",
      "parents": [
        "dbe3168da051766859d9ee6eb8e221719910055f"
      ],
      "author": {
        "name": "vignesh_A",
        "email": "165256748+vigneshio@users.noreply.github.com",
        "time": "Wed Nov 18 19:59:52 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 09:29:52 2026 -0500"
      },
      "message": "fix(nosql): make IndexKey.toSafeString injective for stripe keys (#5872)\n\n* fix(nosql): make IndexKey.toSafeString injective for stripe keys"
    },
    {
      "commit": "dbe3168da051766859d9ee6eb8e221719910055f",
      "tree": "d555c8be81232d3727e8e828c6e212be20893a46",
      "parents": [
        "a40e06aa64fdbd8263644cc03d6bf90dcc06c613"
      ],
      "author": {
        "name": "Ayush Saxena",
        "email": "ayushsaxena@apache.org",
        "time": "Wed Nov 18 17:24:24 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 17:24:24 2026 +0530"
      },
      "message": "Fix a namespace location removal silently moving the namespace (#5876)\n\nRemoving a namespace\u0027s location property does not leave the namespace without a\nlocation. The location is then derived from the namespace name, so the removal\nmoves any namespace that was stored somewhere else, and removeProperties applied\nthat move with no validation at all: it reads the entity, drops the requested\nkeys and writes it back. Its sibling setProperties guards the same mutation\ntwice, through the overlap check when the location changes and through the\ndefault-location check when the payload carries the property.\n\nA namespace stored at a custom location was therefore moved away from the\nlocation its tables were created under, so locations derived for new tables and\ncredentials vended for the namespace pointed somewhere the data is not.\nALLOW_NAMESPACE_CUSTOM_LOCATION governs whether such a location can be set, not\nwhether its removal is checked, so a location stored while the flag was enabled\nmoves just as silently once the flag is off. The property is not filtered out on\nthe way in either, because the reserved-property filter matches reserved\nprefixes and the key is the plain Iceberg name \u0027location\u0027.\n\nThe removal is now rejected when it would change the location the namespace\nresolves to, which leaves the caller to state the new location through a\nproperty update, where it is validated. Removing the property from a namespace\nthat already sits at its name-derived location resolves to the same location\nbefore and after and stays allowed, which covers every namespace created while\ncustom locations are disabled."
    },
    {
      "commit": "a40e06aa64fdbd8263644cc03d6bf90dcc06c613",
      "tree": "4d6b1b64c11e42d32f7d4889b9c46f7151b798ed",
      "parents": [
        "2dd2f1d14d59ec9800f1fb1eee47fe5e91e2b7ae"
      ],
      "author": {
        "name": "Alexandre Dutra",
        "email": "adutra@apache.org",
        "time": "Wed Nov 18 12:27:56 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 12:27:56 2026 +0100"
      },
      "message": "Make TestCacheInvalidationSender.transportFailureMetrics independent of DNS (#6121)\n\nThe test relied on an unresolvable hostname failing fast. On machines\nwhere DNS resolution hangs, the request timeout fired first and the\noutcome was recorded as a timeout instead of a transport error. Use a\nclosed local port instead, which fails immediately with connection\nrefused."
    },
    {
      "commit": "2dd2f1d14d59ec9800f1fb1eee47fe5e91e2b7ae",
      "tree": "cab7322a59fe006f43a7435db7a40b6bc3f6d2fa",
      "parents": [
        "6b9592397415b8a234bbb493824b6b3b6be0160b"
      ],
      "author": {
        "name": "Faris Basha TM",
        "email": "72191505+farisbasha@users.noreply.github.com",
        "time": "Wed Nov 18 14:59:56 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 10:29:56 2026 +0100"
      },
      "message": "docs: Fix dead getting-started and auth source links (#6124)\n\nThe getting-started examples moved from getting-started/ to the\nwebsite guides in #3550, and InternalPolarisToken / TokenBroker moved\nto the auth/internal/broker package. Point the links at the new\nlocations."
    },
    {
      "commit": "6b9592397415b8a234bbb493824b6b3b6be0160b",
      "tree": "f6946980ac98c4eb7d539e4f86b5dff655b82d2c",
      "parents": [
        "ecae57895640839ae692649e1434d3570d4203cc"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Wed Nov 18 08:43:14 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 09:43:14 2026 +0100"
      },
      "message": "Update dependency org.lz4:lz4-java to v1.8.1 (#6125)"
    },
    {
      "commit": "ecae57895640839ae692649e1434d3570d4203cc",
      "tree": "fa7e0155aa974137e343d710d134c38c5b9fbc28",
      "parents": [
        "5a4a487caabcec960fe8a4bcfa598513a5ba64a9"
      ],
      "author": {
        "name": "Akash R Nilugal",
        "email": "akashnilugal@gmail.com",
        "time": "Wed Nov 18 13:09:59 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Wed Nov 18 01:39:59 2026 -0600"
      },
      "message": "Docs: add Kafka, DynamoDB and idempotency to the configuration reference (#6062)\n\nThe configuration reference is assembled from generated files in\nconfiguration/config-sections/, each pulled in with an include-config-section\nline. Three generated files existed but were never included, so their\nproperties were missing from the published reference:\n\n- smallrye-polaris_event_listener_kafka (polaris.event-listener.kafka.*)\n- smallrye-polaris_persistence_nosql_dynamodb\n  (polaris.persistence.nosql.dynamodb.*)\n- smallrye-polaris_idempotency (polaris.idempotency.*)\n\nevents.md already lists Kafka as a supported listener, and MongoDB, the sibling\nof DynamoDB, is already included, but users could not find the settings for\neither. Add the three missing sections next to their related ones, following\nthe existing CloudWatch pattern.\n\nsmallrye-polaris_reserved_properties is intentionally left out: its generated\nfile has a description but no property table."
    },
    {
      "commit": "5a4a487caabcec960fe8a4bcfa598513a5ba64a9",
      "tree": "c15d63b8fa749a8036e5623b94ed5beca454a446",
      "parents": [
        "9102acf1121e2e65000fed9558c3dc5f28a5036e"
      ],
      "author": {
        "name": "youngrae kim",
        "email": "160102021+youngrae81@users.noreply.github.com",
        "time": "Wed Nov 18 16:32:21 2026 +0900"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 23:32:21 2026 -0800"
      },
      "message": "Make `uri` optional for HADOOP connection configurations (#5999)"
    },
    {
      "commit": "9102acf1121e2e65000fed9558c3dc5f28a5036e",
      "tree": "ea38aa043673d1ebfbf1d39d6ec64397c5467f79",
      "parents": [
        "0778b27983f778a69b100ba3b44edd5321c7d5c2"
      ],
      "author": {
        "name": "iting0321",
        "email": "137410100+iting0321@users.noreply.github.com",
        "time": "Wed Nov 18 15:26:17 2026 +0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 23:26:17 2026 -0800"
      },
      "message": "Add Spark integration tests for Apache Paimon tables (#5834)"
    },
    {
      "commit": "0778b27983f778a69b100ba3b44edd5321c7d5c2",
      "tree": "9e94739c9f1e5e7e2daa2fa447d54874e61cc21c",
      "parents": [
        "c8872cb8f332b6a5f13d74e776dcdfa00066cfdd"
      ],
      "author": {
        "name": "Akash R Nilugal",
        "email": "akashnilugal@gmail.com",
        "time": "Wed Nov 18 05:35:57 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 16:05:57 2026 -0800"
      },
      "message": "Allow s3:AbortMultipartUpload in vended AWS write credentials (#6064)"
    },
    {
      "commit": "c8872cb8f332b6a5f13d74e776dcdfa00066cfdd",
      "tree": "f1d2eda927f6895b5f2496d5dd2cdbffacb1b480",
      "parents": [
        "06057cada818dea16307b3f87cb960f5a1696fde"
      ],
      "author": {
        "name": "Alejandro Pajaron",
        "email": "apjorquera@gmail.com",
        "time": "Wed Nov 18 01:03:50 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 19:03:50 2026 -0500"
      },
      "message": "Add Spark plugin integration tests for UPDATE, DELETE and MERGE (#6117)\n\nThe Spark plugin integration tests only covered table creation, inserts,\nreads and schema changes. Nothing checked that the row-level SQL\nstatements work through the Polaris SparkCatalog.\n\nAdd a test for each format the plugin routes:\n- Iceberg (SparkIT): UPDATE, DELETE and MERGE INTO\n- Hudi (SparkHudiIT): UPDATE, DELETE and MERGE INTO on a table with a\n  primary key\n- Delta (SparkDeltaIT): DELETE and MERGE INTO\n\nDelta UPDATE is not covered. When IcebergSparkSessionExtensions is\nloaded, Iceberg\u0027s row lineage rewrite rule fails with a scala.MatchError\non non-Iceberg V2 tables. This also happens for Delta tables in\nspark_catalog, so it is not specific to Polaris."
    },
    {
      "commit": "06057cada818dea16307b3f87cb960f5a1696fde",
      "tree": "fbca2dc1986243508b3457c216f83725dfd90948",
      "parents": [
        "14efae883632e2bf898cb8c5d2a09cde3762ddd3"
      ],
      "author": {
        "name": "Yuewei Zhou",
        "email": "91304878+ZephyrYWZhou@users.noreply.github.com",
        "time": "Tue Nov 17 15:50:01 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 15:50:01 2026 -0800"
      },
      "message": "Persistence: DynamoDB (atomic) write child records in the base table (#6118)"
    },
    {
      "commit": "14efae883632e2bf898cb8c5d2a09cde3762ddd3",
      "tree": "da45f83776f9965c6f15ec9eb82ff1b6c612e735",
      "parents": [
        "0ff3c63631a03356dc470bc5c5380b3250e29cdc"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Tue Nov 17 23:26:41 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 15:26:41 2026 -0800"
      },
      "message": "Update dependency com.google.cloud:google-cloud-iamcredentials to v2.103.0 (#6122)"
    },
    {
      "commit": "0ff3c63631a03356dc470bc5c5380b3250e29cdc",
      "tree": "3533332165e75bb9a1978aefffdff9c3d45e96b0",
      "parents": [
        "417da831e90baf26d6e21df3ea9e4de56d8a310d"
      ],
      "author": {
        "name": "youngrae kim",
        "email": "160102021+youngrae81@users.noreply.github.com",
        "time": "Wed Nov 18 07:34:29 2026 +0900"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 17:34:29 2026 -0500"
      },
      "message": "Let the storage integration decide whether to return the credential refresh endpoint (#6055)\n\n* Let the storage integration decide whether to return the credential refresh endpoint\n\nThe Iceberg REST adapter decided whether to pass the credential refresh endpoint\ndown based on the X-Iceberg-Access-Delegation header, and the storage integration\nadded it to the returned configuration regardless of whether it had vended any\ncredentials. The endpoint is only meaningful when credentials were vended, and the\nstorage integration is the component that knows that.\n\nThe adapter now always computes the (static) credentials path. The handler offers\nit to the storage integration only when the request resolved to credential vending,\nand CachingStorageIntegration adds it to the configuration only when the backend\nreturned credentials. A storage configuration that does not vend credentials\ntherefore no longer advertises an endpoint the client cannot use. Behavior for\nrequests that vend credentials is unchanged.\n\nFixes #6023\n\n* Return the credentials path as a String in the adapter and trim the CHANGELOG entry\n\n* Move the CHANGELOG entry to a changelog.d fragment"
    },
    {
      "commit": "417da831e90baf26d6e21df3ea9e4de56d8a310d",
      "tree": "75284510d3ed022d806798c7036d45880649d5b0",
      "parents": [
        "4b8182a4ee13af5e0666899d025ce180505972d0"
      ],
      "author": {
        "name": "bardrop",
        "email": "mradulsingh524.ms@gmail.com",
        "time": "Wed Nov 18 03:37:30 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 17:07:30 2026 -0500"
      },
      "message": "Reuse one default AWS credentials provider for STS credential vending (#6091)\n\n* Reuse one default AWS credentials provider for STS credential vending\n\nWithout static AWS keys, StorageConfiguration.stsCredentials() builds a\nnew DefaultCredentialsProvider on each call, and StsCredentialVendingMechanism\ncalled it for every AssumeRole request. Each new provider fetches the base\ncredentials again from their source (IMDS, STS web identity, ...), so a burst\nof credential-cache misses becomes a burst of requests to that source.\n\nThe mechanism is application scoped, so resolve the default provider once and\nshare it. Named storages keep their per-name static credentials.\n\nPart of #6090\n\n* Reuse one credentials provider per named storage as well\n\n* Shorten memoize comment and test that unknown storage names are not cached\n\n* Build the shared credentials providers once in the constructor\n\nThe bean is application scoped, so its constructor runs once. Keep the\ndefault and named providers in final fields instead of a memoized\nsupplier."
    },
    {
      "commit": "4b8182a4ee13af5e0666899d025ce180505972d0",
      "tree": "cf5365985f7de2688cccb980df34432390c5d165",
      "parents": [
        "5c62e9379045e42d35cf54bf8d0169fe30f157d0"
      ],
      "author": {
        "name": "Arnab Karmakar",
        "email": "117086972+ArnabKarmakar1108@users.noreply.github.com",
        "time": "Tue Nov 17 20:42:36 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 10:12:36 2026 -0500"
      },
      "message": "Add optional STS endpoint to SigV4 connection authentication (#6035)\n\n* Add optional STS endpoint to SigV4 connection authentication\n\nSigV4 connections always called AWS STS AssumeRole against the SDK\ndefault endpoint, so deployments that need a custom STS endpoint could\nnot use SigV4 federated connections.\n\nAdd an optional stsEndpoint to SigV4AuthenticationParameters. When set,\nSigV4ConnectionCredentialVendor builds its StsDestination with that\nendpoint. When absent, behaviour is unchanged.\n\n* Address review comments on SigV4 STS endpoint validation\n\n* Use example.com test URLs consistent with the OpenAPI spec and remove\nthe reference to ALLOW_SETTING_S3_ENDPOINTS from the STS\nendpoint description.\n\n* Allow http SigV4 STS endpoints with warning and relax URI restrictions"
    },
    {
      "commit": "5c62e9379045e42d35cf54bf8d0169fe30f157d0",
      "tree": "70bf1b1ecab72a37f9431a00bf070feccee0ec9c",
      "parents": [
        "0ab01a464a689346c38bc36f9cecffb81f8ae4bb"
      ],
      "author": {
        "name": "Ayush Saxena",
        "email": "ayushsaxena@apache.org",
        "time": "Tue Nov 17 18:15:53 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 18:15:53 2026 +0530"
      },
      "message": "Fix bootstrap reporting success when the credentials file names no realms (#6114)\n\nbootstrap takes the realms to create from the entries of --credentials-file. A\nfile that parses but carries no entries - an empty file, {}, a YAML document\nthat is only a license header or comments, a secret that was mounted empty -\nyields an empty realm set, so bootstrapRealms iterates nothing and returns an\nempty map.\n\nThe command decides whether it succeeded from that map alone, so an empty map\nis indistinguishable from a run in which every realm was bootstrapped. It\nprints \"Bootstrap completed successfully.\" and exits 0 having created nothing.\nAn automated bootstrap gated on that exit code then proceeds against realms\nthat do not exist, and the failure surfaces later as an authentication error\nwith no obvious cause.\n\nThe standard input path already refuses to bootstrap a realm whose root\nprincipal would be unreachable, rather than reporting a success it cannot\nstand behind. The file path now gets the same treatment: a credentials file\nthat names no realm is reported with its path, and the command exits without\nbootstrapping."
    },
    {
      "commit": "0ab01a464a689346c38bc36f9cecffb81f8ae4bb",
      "tree": "7d54b75dd76811485817a2204c7db2178e144440",
      "parents": [
        "dc8eb486d0c983067369cae2fabf4c38a2af1f59"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Tue Nov 17 09:30:45 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Tue Nov 17 09:30:45 2026 +0100"
      },
      "message": "NoSQL: Fix `IndexKey` hash caching under concurrent access (#6116)\n\nThe separate, non-volatile `hasHash` and `hash` fields allow another thread to observe `hasHash \u003d\u003d true` while still reading the initial zero hash, potentially returning an incorrect hash code.\n\nUse zero as the uncached sentinel and read, compute, and return the hash through a local variable.\nThis avoids a second field read that could observe a stale zero after the first read observed a nonzero hash.\n\nConcurrent calls may redundantly compute the same hash, so no volatile access or synchronization is needed.\nKeys whose actual hash is zero are recomputed on each call, preserving correctness at the cost of caching that case."
    },
    {
      "commit": "dc8eb486d0c983067369cae2fabf4c38a2af1f59",
      "tree": "4c14c329d4c798216715bd5ee22b09d9cca2251e",
      "parents": [
        "0e408fb834af487fe45bd53d471e72f79a2cf39d"
      ],
      "author": {
        "name": "Alexandre Dutra",
        "email": "adutra@apache.org",
        "time": "Mon Nov 16 22:02:59 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Nov 16 22:02:59 2026 +0100"
      },
      "message": "Add changelog.d fragments to avoid CHANGELOG.md merge conflicts (#5967)\n\nAlmost every user-facing PR adds an entry under \"## [Unreleased]\" in\nCHANGELOG.md, so those PRs regularly conflict with each other even though\nthere is no logical conflict. As discussed on the dev mailing list, PRs now add \ntheir changelog entries as small files under changelog.d/, and the release \nmerges them into CHANGELOG.md.\n\nA fragment is changelog.d/\u003cshort-title\u003e.md. It looks like a miniature\nchangelog: a \"### \u003csection\u003e\" heading for each part of the changelog that the\nchange needs an entry in, followed by \"- \" list items. Fragments have no\nlicense header and are excluded from the RAT audit, as the release deletes\nthem."
    },
    {
      "commit": "0e408fb834af487fe45bd53d471e72f79a2cf39d",
      "tree": "51c9d43f5b28abe788434f3632501d6141bffe0c",
      "parents": [
        "cc552849d628098584c378ef4b7826fbed78237d"
      ],
      "author": {
        "name": "vignesh_A",
        "email": "165256748+vigneshio@users.noreply.github.com",
        "time": "Tue Nov 17 00:37:05 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Nov 16 14:07:05 2026 -0500"
      },
      "message": "Add Ranger catalog-config-read for Iceberg REST GET /v1/config (#6011)\n\n* Add Ranger catalog-config-read for Iceberg REST GET /v1/config\n\nRanger maps GET_CATALOG_CONFIG to a dedicated catalog-config-read access\ntype. Existing catalog-level catalog-properties-read and stronger catalog\naccess types imply it. Operators must re-import polaris-ranger-servicedef.json.\n\nCatalog-level content grants still cover GET /v1/config under\ninternal RBAC only. Ranger honors the listed catalog-* types.\n\n* Drop Ranger production-readiness warning for catalog-config-read\n\nRanger Admin is an external system. CHANGELOG and README PUT cover\nthe re-import. A Polaris upgrade does not update Ranger Admin.\n\n* Keep the 1.9.0 CATALOG_READ_CONFIG changelog note unchanged\n\nThe GET /v1/config upgrade note shipped in 1.9.0-rc1. Document the\ndedicated Ranger catalog-config-read access type as a new Unreleased\nentry instead of rewriting that 1.9.0 bullet."
    },
    {
      "commit": "cc552849d628098584c378ef4b7826fbed78237d",
      "tree": "71a47faa09aa0a03da5c179d4d89951fee562c39",
      "parents": [
        "82e5e4a706b934faf442874c5d44402329438509"
      ],
      "author": {
        "name": "vignesh_A",
        "email": "165256748+vigneshio@users.noreply.github.com",
        "time": "Mon Nov 16 16:37:05 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Nov 16 12:07:05 2026 +0100"
      },
      "message": "Fix MemorySize.toString() for 1024 YiB and larger (#6111)\n\nvalueOf(\"1024Y\") is accepted, but toString() kept dividing\nby 1024 past the 9-entry suffix array."
    },
    {
      "commit": "82e5e4a706b934faf442874c5d44402329438509",
      "tree": "48ca9e983fcd0058f83eba81d44151b92ebd4605",
      "parents": [
        "cf0a38448fdaf90cf7a8c2f970693d9b300d4214"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Mon Nov 16 10:44:25 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Nov 16 10:44:25 2026 +0100"
      },
      "message": "Add benchmarks for NoSQL object serialization (#6076)\n\nAdd JMH coverage for cache reconstruction and persistence serialization and deserialization using representative objects and payload sizes, see PRs #6073, #6074, #6075.\n\nValidate fixture types, values and metadata during setup.\nMerge service providers in the benchmark jar so concrete object types resolve correctly while normal module publication remains unchanged."
    },
    {
      "commit": "cf0a38448fdaf90cf7a8c2f970693d9b300d4214",
      "tree": "354aee3fa2aab6e4b94480f7d75437e9d18eaaae",
      "parents": [
        "5d376bdf61f6be1d13bb857f4618ae2bb008a3b8"
      ],
      "author": {
        "name": "Prithvi S",
        "email": "prithvisivasankar@gmail.com",
        "time": "Mon Nov 16 15:13:41 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Nov 16 10:43:41 2026 +0100"
      },
      "message": "NoSQL: honour the requested entity subtype in readEntityByName (#6102)\n\nreadEntityByName ignored entitySubType and returned a TABLE_LIKE\nentity of the wrong subtype. Match Atomic and Transactional: exact\nsubtype unless the caller passes ANY_SUBTYPE.\n\nSigned-off-by: Prithvi S \u003cprithvisivasankar@gmail.com\u003e"
    },
    {
      "commit": "5d376bdf61f6be1d13bb857f4618ae2bb008a3b8",
      "tree": "002643f15313c22c26f4f9382e3e7b04cbc882b9",
      "parents": [
        "a52f81215f836640311858c0c27afec69d8a4c6b"
      ],
      "author": {
        "name": "Prithvi S",
        "email": "prithvisivasankar@gmail.com",
        "time": "Mon Nov 16 14:41:40 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Mon Nov 16 10:11:40 2026 +0100"
      },
      "message": "Reject snowflake timestamps outside the constructId range (#6110)\n\nconstructId masked the timestamp before comparing it, so only the\nreserved system-id value was rejected. Values at or above the bit\nwidth, and most negatives, packed into a garbage id.\n\nSigned-off-by: Prithvi S \u003cprithvisivasankar@gmail.com\u003e"
    },
    {
      "commit": "a52f81215f836640311858c0c27afec69d8a4c6b",
      "tree": "72aa8d053b1e1d536321525ddd41cc78af9dbb58",
      "parents": [
        "b565fe8c7fb273dbb12cae464fcb1758b09a6066"
      ],
      "author": {
        "name": "Ruobing Wang",
        "email": "80226058+Ruobing1997@users.noreply.github.com",
        "time": "Sun Nov 15 12:23:08 2026 -0800"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 12:23:08 2026 -0800"
      },
      "message": "Reject overlapping Generic Table locations on create (#5964)"
    },
    {
      "commit": "b565fe8c7fb273dbb12cae464fcb1758b09a6066",
      "tree": "c6d47beff0017340f4fee27905553f052bc5e729",
      "parents": [
        "4ad26bdc69af51b4127fe4ddd0c2a8d2f2ebd93c"
      ],
      "author": {
        "name": "Alejandro Pajaron",
        "email": "apjorquera@gmail.com",
        "time": "Sun Nov 15 21:22:32 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 12:22:32 2026 -0800"
      },
      "message": "docs: Fix malformed Docsy theme link in website README (#6107)"
    },
    {
      "commit": "4ad26bdc69af51b4127fe4ddd0c2a8d2f2ebd93c",
      "tree": "c2e72641c8b57d7bbf5ce4ad8bce196021574718",
      "parents": [
        "cecf5cf70934ad5e7f031d54f4fce564848712db"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Sun Nov 15 20:12:08 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 21:12:08 2026 +0100"
      },
      "message": "Update docker.io/floci/floci-gcp Docker tag to v0.13.0 (#6109)"
    },
    {
      "commit": "cecf5cf70934ad5e7f031d54f4fce564848712db",
      "tree": "5220d2d052cff9a5024f806eeb1543261a257303",
      "parents": [
        "1b15a6f7e8c512cd0c70db60af8568faa4b5d747"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Sun Nov 15 20:11:56 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 21:11:56 2026 +0100"
      },
      "message": "Update docker.io/floci/floci-az Docker tag to v0.17.0 (#6108)"
    },
    {
      "commit": "1b15a6f7e8c512cd0c70db60af8568faa4b5d747",
      "tree": "75798cf990eb41295d174640d5890c669c400942",
      "parents": [
        "6379d6455fd7980a62e697be6bb614affae3fafe"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Sun Nov 15 13:19:17 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 14:19:17 2026 +0100"
      },
      "message": "Update docker.io/floci/floci Docker tag to v2.5.0 (#6106)"
    },
    {
      "commit": "6379d6455fd7980a62e697be6bb614affae3fafe",
      "tree": "e11b9975d8dcf83a6e47a75d4299eecb9886641a",
      "parents": [
        "934ac8728cde4ec1a17b90c771dcca3876073358"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sun Nov 15 08:41:13 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 08:41:13 2026 +0100"
      },
      "message": "NoSQL: Avoid eagerly materializing index keys during embedded index cleanup (#6100)\n\nOnly retrieve the key when processing a removal marker, where it is needed to check the reference index.\nNon-null entries are retained directly, so accessing their keys unnecessarily triggers lazy key reconstruction and deserialization, adding CPU work and allocations."
    },
    {
      "commit": "934ac8728cde4ec1a17b90c771dcca3876073358",
      "tree": "895c2d0b78a181fa4f8dbf858347f06a0ef1d8d9",
      "parents": [
        "be433613659caf9e988cd219f2151cb1a988478f"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sun Nov 15 08:30:58 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 08:30:58 2026 +0100"
      },
      "message": "Reduce allocation when reconstructing cached NoSQL objects (#6074)\n\nReading a cached NoSQL object decodes its type ID from cached metadata into a new String, although the cache lookup already supplies that ID.\n\nResolve the object type using the lookup’s type ID instead, avoiding that String allocation without increasing the cache entry’s retained weight."
    },
    {
      "commit": "be433613659caf9e988cd219f2151cb1a988478f",
      "tree": "3a72466fcf087c86d323a7dd7ea8455e66b26963",
      "parents": [
        "a70f517a7ab454de71999c669d3e766295411600"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sun Nov 15 08:24:25 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 08:24:25 2026 +0100"
      },
      "message": "NoSQL: Avoid fetching unrelated index stripes during iteration (#6093)\n\nForward prefix iteration prefetched every remaining stripe, and both iteration directions could fetch a stripe outside the requested range when a bound fell between stripes.\nQueries outside the index could also trigger unnecessary loads.\n\nTrim the first and last stripes before prefetching or traversing them, and return an empty iterator when no stripe overlaps.\nLocate the upper stripe using binary search and use current bounds for loaded stripes so in-memory additions and removals remain visible.\n\nAdd regression coverage for prefix and range queries, gaps, open bounds, empty results, modified stripes, both directions, and prefetch enabled or disabled."
    },
    {
      "commit": "a70f517a7ab454de71999c669d3e766295411600",
      "tree": "112d1ca5f06573ce08f89e634f28068ccd82eafe",
      "parents": [
        "c5736fb8781741f6e42adced6f43360308351dae"
      ],
      "author": {
        "name": "JB Onofré",
        "email": "jbonofre@apache.org",
        "time": "Sun Nov 15 06:22:11 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 07:22:11 2026 +0100"
      },
      "message": "Update download page for 1.9.0 release (#6099)"
    },
    {
      "commit": "c5736fb8781741f6e42adced6f43360308351dae",
      "tree": "317d9c1f44c6ee3738e228b74963addf4cf6afa8",
      "parents": [
        "9c4d14a62731429cb991ab389aa7fe7cc3dbd01c"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Sun Nov 15 06:21:01 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 11:51:01 2026 +0530"
      },
      "message": "Update azure/setup-helm action to v5.1.0 (#6032)"
    },
    {
      "commit": "9c4d14a62731429cb991ab389aa7fe7cc3dbd01c",
      "tree": "e07704b852e579ab73f6e0c69aba9efbca3c41fc",
      "parents": [
        "67d10ae1629927fbac048cd4aa09a5906a1cd5dd"
      ],
      "author": {
        "name": "Dmitri Bourlatchkov",
        "email": "dmitri.bourlatchkov@gmail.com",
        "time": "Sun Nov 15 00:31:05 2026 -0500"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sun Nov 15 11:01:05 2026 +0530"
      },
      "message": "Update header on PolarisAdminTool (#6083)\n\n`PolarisAdminTool` code was initially inspired and developed based on its Nessie counterpart: `NessieServerAdminTool`.\n\nHowever, at this time the code similarity is below 50%. Relevant code is specific to the Polaris project.\n\nFunctionally this class is mostly a collection of annotation-based references to sub-commands, and the Polaris sub-commands are materially different from Nessie Admin commands.\n\nCodex Sol (xhigh) analysis:\n\n```\n40.0% exact line similarity, using the same rules as before.\n\n|  | Polaris | Nessie |\n|---|---:|---:|\n| Code lines counted | 5 | 10 |\n| Identical lines, matched in order | 3 | 3 |\n| Different or unmatched lines | 2 | 7 |\n\nThe three matching lines are:\n\npublic Integer call() {\nPrintWriter out \u003d spec.commandLine().getOut();\nreturn 0;\n\nCalculation: 2 × 3 / (5 + 10) \u003d 40.0%.\n\nThis excludes comments, blank lines, imports/packages, complete annotations—including multiline @Command(...) configuration—and brace-only lines. Declarations and method signatures count; indentation is ignored.\n```\n\nThe base class (`BaseCommand`) is completely different in Polaris and Nessie:\n\nCodex Sol (xhigh) analysis:\n\n```\n5.3% exact line similarity, using the same comparison rules.\n\n|  | Polaris | Nessie |\n|---|---:|---:|\n| Code lines counted | 6 | 32 |\n| Identical lines, matched in order | 1 | 1 |\n| Different or unmatched lines | 5 | 31 |\n\nThe only identical line is the class declaration, at Polaris line 25 (.../polaris/runtime/admin/src/main/java/org/apache/polaris/admintool/BaseCommand.java:25) and Nessie line 33 (.../nessie/tools/server-admin/src/main/java/org/projectnessie/tools/admin/cli/BaseCommand.java:33):\n\npublic abstract class BaseCommand implements Callable\u003cInteger\u003e {\n\nCalculation: 2 × 1 / (6 + 32) \u003d 5.26%.\n\nComments, imports/packages, annotations, blank lines, and brace-only lines are excluded. Declarations sharing a line with an annotation are retained; indentation is ignored.\n\nPolaris contains four exit-code constants and a protected spec field. Nessie has different constants, additional injected fields, and methods for warnings and reference/hash resolution.\n```"
    },
    {
      "commit": "67d10ae1629927fbac048cd4aa09a5906a1cd5dd",
      "tree": "c2bba581a2a2358d04c4d5c0a2e9012366e33c68",
      "parents": [
        "3ab75b9012883c6853613784249dd7bac049b069"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Sat Nov 14 18:09:01 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 19:09:01 2026 +0100"
      },
      "message": "Update dependency ch.qos.logback:logback-classic to v1.6.7 (#6101)"
    },
    {
      "commit": "3ab75b9012883c6853613784249dd7bac049b069",
      "tree": "6a39ccee9cd15a7142fc709ab962a306999049b7",
      "parents": [
        "a529055716d58f50ba67be765cfb3d0f955cc2ab"
      ],
      "author": {
        "name": "Ayush Saxena",
        "email": "ayushsaxena@apache.org",
        "time": "Sat Nov 14 22:24:05 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 22:24:05 2026 +0530"
      },
      "message": "Fix commitTransaction failing with 500 instead of 400 when a table\u0027s locations change (#5742)\n\nA multi-table transaction stages its writes in TransactionWorkspaceMetaStoreManager,\nwhich refuses listEntities so that a read inside the transaction cannot imply it sees\nthe transaction\u0027s own uncommitted writes. The location-overlap validation needs that\nlisting, so commitTransaction rejects up front any commit that would reach it.\n\nThat guard lived in IcebergCatalogHandler and decided, update by update, what\nLocalIcebergCatalog decides from the base and the final metadata, so the two drifted.\nThe guard recognised only SetLocation, while the catalog also compares the\nwrite.data.path and write.metadata.path properties, so a SetProperties that sets\neither one, or a RemoveProperties that drops one, passed the guard and reached the\noverlap check during the commit. listEntities threw, the diagnostics assert surfaced\nas IllegalStateException, and the request failed with 500 Internal Server Error where\nthe guard beside it intends 400. The single-table updateTable path accepts the same\nupdate, so the two commit endpoints disagreed. A table whose entity does not record\nits locations, as entities written before 1.3.0 do not, reached the same check on any\nupdate at all, even one that leaves the locations alone.\n\nReject inside the validation instead, where the decision is already made, after the\nALLOW_TABLE_LOCATION_OVERLAP short-circuit and the view condition so that it fires\nexactly when the listing would run. Every trigger is then covered by construction,\nincluding any location property added to the catalog\u0027s comparison later, and a\ntransaction whose updates cancel each other out is no longer rejected.\n\nThe guard also read ALLOW_NAMESPACE_LOCATION_OVERLAP, while the check it stood in for\nreads ALLOW_TABLE_LOCATION_OVERLAP with the catalog it applies to. There is now one\nreader of one option. Both default to false, so the default behaviour is unchanged."
    },
    {
      "commit": "a529055716d58f50ba67be765cfb3d0f955cc2ab",
      "tree": "dcf42dcaa5a1321f4d366f0d04c5d56cb7e62ad8",
      "parents": [
        "9776de004f16010d9ffbe82517bc78888975f576"
      ],
      "author": {
        "name": "Akash R Nilugal",
        "email": "akashnilugal@gmail.com",
        "time": "Sat Nov 14 15:31:13 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 04:01:13 2026 -0600"
      },
      "message": "CLI: require --policy-type for `policies create` (#6078)"
    },
    {
      "commit": "9776de004f16010d9ffbe82517bc78888975f576",
      "tree": "cfee6845f6656f9324d55ce1979b39a6999beec2",
      "parents": [
        "71c8ff6deb6a66dd4bbfccd953ee2df102246c76"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sat Nov 14 09:51:21 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 09:51:21 2026 +0100"
      },
      "message": "NoSQL: Avoid building unused indexes during location overlap checks (#6094)\n\nLocation overlap checks can finish without resolving any entity IDs or names.\nEagerly constructing both lookup indexes does unnecessary work in those cases.\n\nMemoize the ID and name index suppliers and access them inside the stream mapping lambdas.\nEach index is built only on its first lookup and reused across subsequent lookups in the same overlap check."
    },
    {
      "commit": "71c8ff6deb6a66dd4bbfccd953ee2df102246c76",
      "tree": "c44aeb872f0caa24e513244261a5ca7b405a0bc6",
      "parents": [
        "b24928582fd30de071244d869e8e31004009a267"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Sat Nov 14 08:23:08 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 08:23:08 2026 +0000"
      },
      "message": "Update dependency helm to v4.3.2 (#6096)"
    },
    {
      "commit": "b24928582fd30de071244d869e8e31004009a267",
      "tree": "9b5f83d45717d8b720e3f4b15afca025e29a1584",
      "parents": [
        "548f28666ef7e60081b5ffcadbd640e707349713"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Sat Nov 14 08:21:00 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 08:21:00 2026 +0000"
      },
      "message": "Update actions/setup-java action to v6.3.0 (#6097)"
    },
    {
      "commit": "548f28666ef7e60081b5ffcadbd640e707349713",
      "tree": "26408f61e2da3e1baaa63a1538487449d8fdd88b",
      "parents": [
        "a32b12352a57f8f0e2511ca8cf608780a042e4f1"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sat Nov 14 08:43:54 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 08:43:54 2026 +0100"
      },
      "message": "Improve release reproducibility and record build environments (#6079)\n\nJDK patch releases can change compiler and Javadoc output, while release verification offers little context for diagnosing differences.\nCached outputs built with another JDK can also obscure reproduction attempts.\n\nTrack selected JDK runtime versions and vendors as inputs to Java/Scala compilation and documentation tasks.\nPin release JDK and Helm versions, with Renovate maintaining their updates.\n\nGenerate standalone build-environment JSON using the existing signing and SHA-512 helpers.\nStage it with release artifacts, attach it to GitHub releases and summarize it in CI.\nCompare relevant fields during release verification while preserving streamed execution and older candidates.\nKeep run-specific evidence outside archives and byte-for-byte comparisons."
    },
    {
      "commit": "a32b12352a57f8f0e2511ca8cf608780a042e4f1",
      "tree": "1ee22068082d84b97b411c9e5faeb80edf41e87e",
      "parents": [
        "94897a6835c9dc7672c9ab4c65b0c22936eec953"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sat Nov 14 08:43:29 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 08:43:29 2026 +0100"
      },
      "message": "Reduce NoSQL persistence deserialization overhead (#6075)\n\nSingle-part objects already arrive as byte arrays, but `fetchMany` wraps them in a stream before parsing.\nDeserialization also constructs typed readers repeatedly for the same mapper and target class.\n\nParse single-part payloads directly from bytes and reuse typed readers for object and generic deserialization.\nCentralize object-reader setup so stream and byte-array inputs apply the same type validation and metadata injection.\n\nMultipart objects continue to use the stream path."
    },
    {
      "commit": "94897a6835c9dc7672c9ab4c65b0c22936eec953",
      "tree": "934076fdc1941bc08d0825be76a8be5a358e44e5",
      "parents": [
        "1d7a52f250d685c6eb880bdef203296c94a63a91"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sat Nov 14 08:43:12 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 08:43:12 2026 +0100"
      },
      "message": "NoSQL: Reduce entity-ID set decoding allocation and resizing (#6072)\n\nThe decoder knows its encoded element count but used an unsized generated set builder, growing and copying temporary storage while reading IDs.\nBuild a presized Guava immutable set and pass it through the existing immutable factory, preserving validation, deduplication, iteration order, and wire bytes.\nBound the initial capacity by count, remaining bytes, and 1024 to avoid large speculative reservations on truncated or duplicate-heavy inputs.\n\nWrite the canonical zero-byte null marker directly.\nKeep the original `null` probe: its one-byte fast-path experiment regressed multi-byte counts.\nAdd JMH coverage for `null`/singleton/medium/large sets and small versus large IDs, plus regression tests for duplicate IDs and a truncated enormous declared count.\n\nJMH on Ryzen 9 7950X / Zulu JDK 21.0.12, one thread, two forks,\n3 warmup + 5 measurement iterations of 400 ms, GC profiler, 256 MiB heap.\nCardinality -1 means null:\n\n| Benchmark / parameters | Before ns/op | After ns/op | Time reduction | Before B/op | After B/op |\n| --- | ---: | ---: | ---: | ---: | ---: |\n| deserialize / -1, LARGE | 1.707 | 1.806 | -5.8% | 0.000 | 0.000 |\n| deserialize / 1, LARGE | 17.431 | 16.320 | 6.4% | 136.000 | 88.000 |\n| deserialize / 16, LARGE | 257.016 | 188.080 | 26.8% | 1064.004 | 728.003 |\n| deserialize / 1024, LARGE | 16921.325 | 13412.419 | 20.7% | 58256.294 | 37040.233 |\n\n| Benchmark / parameters | Before ns/op | After ns/op | Time reduction | Before B/op | After B/op |\n| --- | ---: | ---: | ---: | ---: | ---: |\n| deserialize / 1, SMALL | 11.248 | 10.748 | 4.5% | 112.000 | 64.000 |\n| deserialize / 16, SMALL | 178.735 | 113.497 | 36.5% | 680.003 | 344.002 |\n\n| Benchmark / parameters | Before ns/op | After ns/op | Time reduction | Before B/op | After B/op |\n| --- | ---: | ---: | ---: | ---: | ---: |\n| sizeAndSerialize / -1, LARGE | 1.647 | 1.492 | 9.4% | 0.000 | 0.000 |\n\nThe null decoder is unchanged and varied by ~0.1 ns; no null-decode improvement is claimed.\nLarge-ID decoding removes 48/336/~21216 B per operation at sizes 1/16/1024.\nResults describe local microbenchmarks, not application throughput."
    },
    {
      "commit": "1d7a52f250d685c6eb880bdef203296c94a63a91",
      "tree": "582da3db56476bf615a79aa889effe31c079c994",
      "parents": [
        "e76927a0ca9817cc92ba02369a27aa1cb48109c6"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sat Nov 14 08:42:53 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 08:42:53 2026 +0100"
      },
      "message": "NoSQL: Avoid repeated Smile encoding of fieldless changes (#6071)\n\nIndex sizing fully serialized every change before serializing it again to write.\nNull and generated add/update/remove values have constant payloads, so retain their mapper-generated Smile bytes and reuse them for sizing/writing.\nUse exact generated classes to preserve payloads on generic implementations with known type IDs.\nKeep Jackson for rename and other payload-bearing changes, and initialize common encodings lazily to avoid interface initialization cycles.\n\nRead the common one-byte length prefix and null payload with absolute buffer accesses.\nRetain the existing varint reader for multi-byte/overlong prefixes.\n\nAdd metastore-types JMH support, serializer benchmarks, byte-for-byte mapper\ncomparisons, generic fallback checks, and buffer/overlong-prefix probe tests.\n\nJMH on Ryzen 9 7950X / Zulu JDK 21.0.12, one thread, two forks,\n3 warmup + 5 measurement iterations of 400 ms, GC profiler, 256 MiB heap:\n\n| Benchmark / parameters | Before ns/op | After ns/op | Time reduction | Before B/op | After B/op |\n| --- | ---: | ---: | ---: | ---: | ---: |\n| isNull / NULL | 5.103 | 3.958 | 22.4% | 0.000 | 0.000 |\n| isNull / ADD | 2.680 | 1.583 | 40.9% | 0.000 | 0.000 |\n| isNull / RENAME | 2.705 | 1.592 | 41.1% | 0.000 | 0.000 |\n| isNull / GENERIC | 2.715 | 1.609 | 40.7% | 0.000 | 0.000 |\n| serialize / NULL | 64.991 | 2.488 | 96.2% | 664.001 | 0.000 |\n| serialize / ADD | 105.097 | 3.283 | 96.9% | 744.002 | 0.000 |\n| serialize / RENAME | 114.594 | 126.836 | -10.7% | 856.002 | 832.002 |\n| serialize / GENERIC | 110.467 | 116.673 | -5.6% | 832.002 | 844.002 |\n| size / NULL | 54.466 | 0.638 | 98.8% | 672.001 | 0.000 |\n| size / ADD | 92.761 | 0.673 | 99.3% | 744.002 | 0.000 |\n| size / RENAME | 105.387 | 125.597 | -19.2% | 816.002 | 792.002 |\n| size / GENERIC | 110.918 | 115.189 | -3.9% | 808.002 | 820.002 |\n| sizeAndSerialize / NULL | 126.279 | 3.042 | 97.6% | 1336.002 | 0.000 |\n| sizeAndSerialize / ADD | 184.806 | 3.256 | 98.2% | 1560.003 | 0.000 |\n| sizeAndSerialize / RENAME | 246.627 | 229.430 | 7.0% | 1672.004 | 1672.004 |\n| sizeAndSerialize / GENERIC | 220.463 | 225.935 | -2.5% | 1664.004 | 1664.004 |\n\nFallback rows are noisy controls, not claimed improvements.\nA focused rename repeat with 5 x 1s warmup did not reproduce the initial regression:\n\n| Benchmark / parameters | Before ns/op | After ns/op | Time reduction | Before B/op | After B/op |\n| --- | ---: | ---: | ---: | ---: | ---: |\n| serialize / RENAME | 114.328 | 112.632 | 1.5% | 856.002 | 856.002 |\n| size / RENAME | 130.270 | 114.772 | 11.9% | 768.002 | 792.002 |\n\nThe probe already allocated approximately zero after escape analysis; its gain is CPU only.\nCached null/add size-plus-write removes ~1.3/1.6 KB per operation.\nLocal microbenchmark results do not establish application throughput gains."
    },
    {
      "commit": "e76927a0ca9817cc92ba02369a27aa1cb48109c6",
      "tree": "8640fc4b29fabd592f419c97366371a51322481e",
      "parents": [
        "d0b64c0cd85ebcbceb6bf050ada9f3a951f06b0e"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sat Nov 14 08:42:34 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 08:42:34 2026 +0100"
      },
      "message": "NoSQL: Reduce object-reference decoding allocations (#6070)\n\n* NoSQL: Reduce object-reference decoding allocations\n\nEvery heap-buffer `ObjRef` decode copied its type name into a temporary byte array before constructing the `String`.\nDecode directly from accessible backing arrays, respecting array offsets and logical limits, to remove that copy and allocation.\nKeep the existing copy path for direct and read-only buffers.\n\nAdd JMH coverage for sizing, serialization, size-plus-write, and decoding across short/32-byte ASCII, Unicode, multipart references, and buffer variants.\nCover UTF-8 replacement, sliced offsets, trailing bytes, and truncated logical limits in regression tests.\nExperiments with ASCII sizing and direct-writing experiments were rejected because their long-name and Unicode regressions outweighed the benefits.\n\nJMH on Ryzen 9 7950X / Zulu JDK 21.0.12, one thread, two forks,\n3 warmup + 5 measurement iterations of 400 ms, GC profiler, 256 MiB heap:\n\n| Benchmark / parameters | Before ns/op | After ns/op | Time reduction | Before B/op | After B/op |\n| --- | ---: | ---: | ---: | ---: | ---: |\n| deserialize / HEAP, ASCII | 15.117 | 13.734 | 9.1% | 104.000 | 80.000 |\n| deserialize / HEAP, ASCII32 | 16.754 | 13.962 | 16.7% | 152.000 | 104.000 |\n| deserialize / HEAP, UNICODE | 21.344 | 18.626 | 12.7% | 128.000 | 104.000 |\n| deserialize / HEAP, MULTIPART | 15.643 | 13.122 | 16.1% | 104.000 | 80.000 |\n\nEach heap decode removes 24-48 temporary bytes.\nDirect/read-only decoding keeps its original implementation; these are microbenchmark rather than application throughput results.\n\n* review"
    },
    {
      "commit": "d0b64c0cd85ebcbceb6bf050ada9f3a951f06b0e",
      "tree": "b2a8b493d04119968ad65eade19ee82422c9552f",
      "parents": [
        "c238bf347f501310645e655cf9fbedfe96ea8128"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Sat Nov 14 08:42:17 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Sat Nov 14 08:42:17 2026 +0100"
      },
      "message": "NoSQL: Reduce variable-length integer sizing CPU cost (#6069)\n\nIndex sizing scans entity IDs and framing lengths repeatedly.\nCounting seven-bit groups in a loop makes that cost grow with the encoded width.\nKeep the cheap one-byte path and calculate wider lengths from their significant-bit count.\nEncoding, decoding, skipping, and negative-input rejection remain unchanged.\n\nAdd JMH support and benchmarks for sizing, reading, writing, and skipping small, large, and mixed-width values, plus length-boundary/negative-input tests.\nThe unconditional formula regressed small inputs; the retained hybrid avoids that regression.\nA signed-byte skip experiment regressed large values and was not retained.\n\nJMH on Ryzen 9 7950X / Zulu JDK 21.0.12, one thread, two forks, 3 warmup + 5 measurement iterations of 400 ms, GC profiler, 256 MiB heap.\n\nThe benchmark annotations default to 1 s iterations. To match the reported iteration settings, pass JMH `-wi 3 -w 400ms -i 5 -r 400ms -f 2 -t 1`; also enable `-prof gc` and configure the forked JVM with a 256 MiB heap.\n\n| Benchmark / parameters | Before ns/op | After ns/op | Time reduction | Before B/op | After B/op |\n| --- | ---: | ---: | ---: | ---: | ---: |\n| length / SMALL | 0.594 | 0.590 | 0.7% | 0.000 | 0.000 |\n| length / LARGE | 2.656 | 1.153 | 56.6% | 0.000 | 0.000 |\n| length / MIXED | 1.558 | 1.134 | 27.3% | 0.000 | 0.000 |\n\nSmall-value sizing is unchanged within noise; successful calls remain allocation-free.\nThese are local microbenchmark results, not application gains."
    },
    {
      "commit": "c238bf347f501310645e655cf9fbedfe96ea8128",
      "tree": "348aa5402d1ee3ab69c9d11a469fded68849ba8d",
      "parents": [
        "9c9f06c63262f976c02db10d8fcb0fad1b285c7f"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Fri Nov 13 18:24:39 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 18:24:39 2026 +0100"
      },
      "message": "Replace JK1 license reports with Licensee checks (#6027)\n\nThe JK1 dependency-license-report pipeline is incompatible with Gradle\u0027s configuration cache.\nReplace it ahead of Gradle\u0027s move to enabling the configuration cache by default, originally targeted for Gradle 10.\n\nChoose Licensee for configuration-cache-compatible policy checks over effective Maven POMs, including inherited license metadata.\nIts explicit SPDX, URL and dependency allowances express the shared policy and narrow exceptions without duplicating Maven model parsing in Polaris build logic.\n\nCheck the dependencies actually selected for the admin/server applications and Spark bundles, and verify that delivered LICENSE/NOTICE files match the attribution inputs.\nInclude the binary ZIP/TAR distributions, retain module check and Maven publication gates, and check both applications before release image builds.\nRemove legacy JK1 report generation and aggregation while keeping packaged LICENSE and NOTICE files.\nThe separate SBOM follow-up adds published CycloneDX dependency and license documents.\n\nKeep policy parsing, validation and Licensee configuration in the typed Kotlin plugin under gradle/license-check-plugin, with focused unit and integration tests.\nLoad it through the small applied-script bootstraps with private buildscript classpaths:\nLicensee\u0027s Maven/Plexus dependencies conflict with Quarkus on a shared plugin classpath and cause Maven XML service-loading failures.\nThe included build alone does not provide that runtime isolation."
    },
    {
      "commit": "9c9f06c63262f976c02db10d8fcb0fad1b285c7f",
      "tree": "d194dd92872a7e9729f796003f010be68260f715",
      "parents": [
        "7bcb5193c7df5dd6f67c0a590c59019baf70f334"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Fri Nov 13 18:23:47 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 18:23:47 2026 +0100"
      },
      "message": "NoSQL: Use fair admission for locally synchronized NoSQL commits (#6039)\n\nAn unfair semaphore allows new commits to overtake queued commits, which can exhaust their retry timeout without attempting a write.\nJVM-wide fairness sleeps also delay locally serialized attempts and couple otherwise unrelated references.\n\nUse a fair semaphore per realm and reference, and explicitly bypass fairness sleeps for locally synchronized commits.\nPreserve configured fairness for other retry loops and retain randomized conflict backoff outside the semaphore.\n\nDeduct fairness admission time from the remaining retry budget and report a timeout before invoking an attempt when that budget is exhausted.\nAdd regression coverage for policy selection, backoff, queued admission, permit release, interruption, and timeout accounting. Document the local fairness override."
    },
    {
      "commit": "7bcb5193c7df5dd6f67c0a590c59019baf70f334",
      "tree": "94141886be2e04167cbebd73f1a29ebe181d863e",
      "parents": [
        "644c727ce3cb8d61d915f00ef6f9446f83bae1b6"
      ],
      "author": {
        "name": "vignesh_A",
        "email": "165256748+vigneshio@users.noreply.github.com",
        "time": "Fri Nov 13 21:55:18 2026 +0530"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 11:25:18 2026 -0500"
      },
      "message": "Enable catalog config authorization in Spark federation ITs (#6065)\n\nRelated to #5998"
    },
    {
      "commit": "644c727ce3cb8d61d915f00ef6f9446f83bae1b6",
      "tree": "b5ce9fa5a10364d36a13182f6c605578a9655a71",
      "parents": [
        "4ee02d5d0a1d2945d69741f20d0f8772c95bac11"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Fri Nov 13 16:49:20 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 16:49:20 2026 +0100"
      },
      "message": "Reduce allocation when binding NoSQL object metadata (#6073)\n\nCache and persistence deserialization build a five-entry injection map for every object read.\n\nReplace the map with an immutable field-based provider.\nProvide a mapper-local typed-reader factory and a binding overload that attaches object-specific metadata to a reusable reader.\n\nCover concurrent metadata binding with JSON and Smile mappers, including `null` version tokens."
    },
    {
      "commit": "4ee02d5d0a1d2945d69741f20d0f8772c95bac11",
      "tree": "6fe188488b68713a14475480c5f5e82162de1631",
      "parents": [
        "1b71cc2fc7b5e5843de3425caa6bafd866c56350"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Fri Nov 13 16:47:01 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 16:47:01 2026 +0100"
      },
      "message": "Throttle database reads during NoSQL retained-object identification (#6054)\n\nRetained-object identification bypasses persistence caches and can put significant pressure on the database.\nExisting maintenance rate limits only paced the subsequent backend scan, leaving identification reads unthrottled.\n\nAdd a backend-wrapper option to the realm persistence builder and use a maintenance-local `ThrottledBackend` to pace identification reads with the existing object and reference rate settings.\nShare limiters across realms and identifiers, counting requested references and object parts, including nested index and non-recording cleanup reads.\n\nPreserve backend observation and existing batch sizes.\nAdd regression coverage and document the expanded rate-limit behavior."
    },
    {
      "commit": "1b71cc2fc7b5e5843de3425caa6bafd866c56350",
      "tree": "1bc5ecfdfa925ec015b48ed8131459c36e48265b",
      "parents": [
        "448db25a7a304afd69f0b3ae4d80e1f248fb712d"
      ],
      "author": {
        "name": "Alexandre Dutra",
        "email": "adutra@apache.org",
        "time": "Fri Nov 13 11:02:18 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 10:02:18 2026 +0000"
      },
      "message": "Remove synthetic principals hack in Resolver (#6020)\n\nExternal principals no longer get synthetic principal and principal role\nentities in the Resolver. Authorizers must resolve only the securables they\nneed via AuthorizationState.resolve(). LocalIcebergCatalog sibling location\nresolution now selects only the reference catalog and requested paths.\n\nFixes #5439"
    },
    {
      "commit": "448db25a7a304afd69f0b3ae4d80e1f248fb712d",
      "tree": "ea1254602da2633d930e202537664d66c117b4c0",
      "parents": [
        "44962b9c27b5c90567e5d6ac6e99237b323f1046"
      ],
      "author": {
        "name": "Yong Zheng",
        "email": "yongzheng0809@gmail.com",
        "time": "Fri Nov 13 01:35:23 2026 -0600"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 01:35:23 2026 -0600"
      },
      "message": "spotlessApply (#6010)"
    },
    {
      "commit": "44962b9c27b5c90567e5d6ac6e99237b323f1046",
      "tree": "929b4642a5cbeb018e192d6804d9d4832103c1b5",
      "parents": [
        "4e15832eaf22dd2aa0ec77f7982aedd1764ba335"
      ],
      "author": {
        "name": "JB Onofré",
        "email": "jbonofre@apache.org",
        "time": "Fri Nov 13 03:51:14 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 09:21:14 2026 +0530"
      },
      "message": "Add Polaris Community Meeting 2026-10-29 (#6080)"
    },
    {
      "commit": "4e15832eaf22dd2aa0ec77f7982aedd1764ba335",
      "tree": "3b8c75ad2b5a5a9c37402b67127acc220f6ee46e",
      "parents": [
        "bc6f88ba088e51f0583501dc40758c1113a8b328"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Fri Nov 13 03:46:41 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 09:16:41 2026 +0530"
      },
      "message": "Update registry.access.redhat.com/ubi9/openjdk-21-runtime Docker tag to v1.24-3.1794341129 (#6086)"
    },
    {
      "commit": "bc6f88ba088e51f0583501dc40758c1113a8b328",
      "tree": "85e5d941330b1a7a7c013a4b4d3cbb23493073fc",
      "parents": [
        "5ef50fd887c937b06144bac43a73bbe51152bb43"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Fri Nov 13 03:44:55 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Fri Nov 13 09:14:55 2026 +0530"
      },
      "message": "Update postgres Docker tag to v18.7 (#6087)"
    },
    {
      "commit": "5ef50fd887c937b06144bac43a73bbe51152bb43",
      "tree": "0f4a580cb4a0694ce74715d6b28d7d6907ca5431",
      "parents": [
        "6906e42ee39f2e85af3ee4bb686552f62e98a439"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Thu Nov 12 19:10:32 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 11:10:32 2026 -0800"
      },
      "message": "Update dependency boto3 to ~\u003d1.44.1 (#6050)"
    },
    {
      "commit": "6906e42ee39f2e85af3ee4bb686552f62e98a439",
      "tree": "1562258cc6e3f5a3a21f7c75584cf8d0bcfb6faf",
      "parents": [
        "8aa75cea8fab927de78c5e288fc38fae3847e2f4"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Thu Nov 12 17:01:32 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 17:01:32 2026 +0100"
      },
      "message": "Make supported locales independent of the build environment (#6056)\n\nQuarkus defaults `quarkus.locales` to the build JVM\u0027s language and country and embeds that value in generated bytecode.\nDifferent builder locales therefore produce different `generated-bytecode.jar` files and binary distributions, preventing byte-for-byte reproducible releases.\nSetting `quarkus.default-locale` alone does not override the supported locale set.\n\nSet `quarkus.locales\u003den-US` in the shared runtime defaults, matching the existing English/US default locale, and document why it must be explicit."
    },
    {
      "commit": "8aa75cea8fab927de78c5e288fc38fae3847e2f4",
      "tree": "ea66c63afcb66c532821a9c166e6ba307d5b6a92",
      "parents": [
        "fb74a5077996a38f03452e263c50c99edc3d074f"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Thu Nov 12 15:23:39 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 15:23:39 2026 +0100"
      },
      "message": "NoSQL: Add metrics for distributed cache invalidations (#6038)\n\nAn empty sender queue does not imply successful delivery, and HTTP success can hide receiver parsing failures.\nLogs alone make it difficult to assess invalidation traffic, delivery failures, and discovery health.\n\nAdd counters, timers, and gauges for invalidation volume, skipped entries, batch sizes, queue depth, in-flight requests, delivery and receiver outcomes, and peer discovery.\nUse fixed labels without peer or realm identifiers to keep cardinality bounded.\n\nKeep histograms disabled by default and support an absent `MeterRegistry`.\nPreserve the wire protocol and existing invalidation behavior.\nDocument metric semantics and cover asynchronous completion and error paths."
    },
    {
      "commit": "fb74a5077996a38f03452e263c50c99edc3d074f",
      "tree": "07bd221abb86812ac98eb1fdb8f45a2997972d78",
      "parents": [
        "56b1bcbb878138de0ebf19dcc0017b2c5afa5e19"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Thu Nov 12 14:08:48 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 15:08:48 2026 +0100"
      },
      "message": "Update openpolicyagent/opa Docker tag to v1.22.4 (#6068)"
    },
    {
      "commit": "56b1bcbb878138de0ebf19dcc0017b2c5afa5e19",
      "tree": "f00de06153845eacf1f9fa46d6b5ad35726bff95",
      "parents": [
        "509e850b9804f8f91869027ed844ad97b0777ca0"
      ],
      "author": {
        "name": "Mend Renovate",
        "email": "bot@renovateapp.com",
        "time": "Thu Nov 12 14:08:44 2026 +0000"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 15:08:44 2026 +0100"
      },
      "message": "Update docker.io/openpolicyagent/opa Docker tag to v1.22.4 (#6067)"
    },
    {
      "commit": "509e850b9804f8f91869027ed844ad97b0777ca0",
      "tree": "ce9ef769d52544c8993d5e61e2beebe41c508d29",
      "parents": [
        "4b80131936b0ef5680bddb608d3758f90d916d7e"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Thu Nov 12 14:14:38 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 14:14:38 2026 +0100"
      },
      "message": "NoSQL: Reduce serialization overhead in the NoSQL Caffeine cache (#6049)\n\nSerialized caching limits retained object graphs and keeps weight accounting predictable, but Smile round trips add unnecessary work for small references and index stripes containing an already-binary payload.\n\nCache immutable `Reference` snapshots directly and retain `IndexStripeObj` payloads as owned byte arrays.\nSupport array-backed, read-only and direct buffers while preserving metadata and independent writable results.\nRecalculate retained weight on every insertion and replacement.\n\nDisable seven-bit binary encoding for remaining cache Smile values, avoiding its roughly 14% binary expansion and associated encoding/decoding work.\nEncode `Instant` and `Duration` as integer seconds/nanoseconds pairs, preserving precision and range while avoiding text and decimal conversion.\nKeep these encoding changes local to the cache.\n\nAdd JMH benchmarks and regression coverage for ownership, replacement weights, metadata and serialization boundaries.\n\nRepresentative local JMH results, Java 21, two forks, one thread:\n- Reference hit, 20-history: 529 -\u003e 46 ns/op; 2220 -\u003e 88 B/op\n- Stripe hit, 64 KiB heap buffer: 17.7 -\u003e 3.7 us/op\n- Stripe update, 64 KiB: 26.4 -\u003e 3.7 us/op;\n  allocation 137528 -\u003e 65720 B/op\n- Binary decode, 64 KiB: 18.2 -\u003e 4.4 us/op\n- Instant decode: 504 -\u003e 94 ns/op; 2328 -\u003e 688 B/op\n- Duration decode: 227 -\u003e 92 ns/op; 947 -\u003e 688 B/op\n\nDirect references trade slightly higher estimated retained weight for lower CPU and allocation: the representative 20-history entry grows from 470 to 656 bytes.\nResults measure individual operations, not end-to-end application performance."
    },
    {
      "commit": "4b80131936b0ef5680bddb608d3758f90d916d7e",
      "tree": "f52408a46b81ee0505ea654a5dc5be0b4054b75b",
      "parents": [
        "5af1634709f685524e7b2e5b64f826c3b4ce222d"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Thu Nov 12 14:14:17 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 14:14:17 2026 +0100"
      },
      "message": "NoSQL: Reduce allocation and buffer overhead in NoSQL index keys (#6047)\n\n`IndexKey` decoding, skipping, and serialization are hot paths.\nAll production call sites use `byte[]`-backed `ByteBuffer`s: serialization and prefix reconstruction allocate heap buffers, and Jackson wraps decoded binary values in heap buffers.\n\nAdd array fast paths using local cursors, accounting for array offsets, positions, and limits.\nRetain generic-buffer fallbacks for direct buffers and read-only heap buffers, which do not expose their backing arrays.\nPreserve exact consumption, exception positions, and partial writes, including truncation halfway through an escape pair.\n\nNamespace/name keys contain UTF-8 elements separated by byte zero.\nNeither the separators nor UTF-8 multibyte sequences require escaping.\nOrdinary identifiers can therefore be scanned and copied directly into an exactly sized result array, avoiding the previous 500-byte scratch array and intermediate writes.\nNull decoding also avoids scratch allocation.\n\nLong IDs and composite binary keys can require escaping.\nKeep the escape-aware decoder, but bound its scratch allocation using the decoded prefix and remaining wire bytes, capped at `MAX_LENGTH`.\nThis particularly benefits short escaped keys and right-sized reconstruction buffers without introducing per-key scratch-pool acquisition.\n\nCheck decoded length before writing scratch so overlong keys produce the intended `IllegalArgumentException`.\nUse an upfront length check and `Arrays.equals` over ranges for `startsWith`.\n\nAdd API-module JMH support and benchmarks covering names, namespaces, plain and escaped longs, composite keys, nulls, and prefix matching.\n\nRepresentative before/after JMH means for sliced heap buffers:\n```\n                              ns/op             B/op\nOperation                  before  after     before  after\nDecode name                 35.95  12.18        568     48\nDecode namespace            45.67  28.72        600     80\nDecode escaped long         39.46  16.78        568     80\nSkip namespace              26.00  13.56          0      0\nSkip escaped long           10.01   6.61          0      0\nSerialize namespace         19.39  17.17          0      0\nSerialize partial namespace 13.21  10.70          0      0\nSerialize partial name       5.48   6.36          0      0\n```\n\nHeap null decoding drops from 27.68 to 1.72 ns/op and from 520 to 0 B/op.\nLong prefix matches improve from 11.22 to 5.86 ns/op; short prefix matches increase from 4.06 to 4.79 ns/op.\nShort-key costs remain a tradeoff rather than an across-the-board speedup.\n\nExplicit-shift and VarHandle long-conversion alternatives showed no clear overall benefit, so retain the existing key(long)/asLong methods."
    },
    {
      "commit": "5af1634709f685524e7b2e5b64f826c3b4ce222d",
      "tree": "8183932f891606f5c965f2173b8fe1d5a8f85dfb",
      "parents": [
        "6c1b361daf3e36294363812ad378d7e052da1878"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Thu Nov 12 14:13:53 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 14:13:53 2026 +0100"
      },
      "message": "NoSQL: Defer construction of write indexes until first access (#6045)\n\nCatalog commits construct write views for indexes that callbacks may never access, unnecessarily deserializing embedded entries and allocating stripe metadata.\n\nReturn lazy write views for existing containers and reuse the original container when finalized without access, unless spilling is required.\nKeep fresh indexes eager and preserve the existing API.\n\nAdapt tests that inspect index internals and add coverage for deferred construction, concurrent initialization, mutation, spilling, and finalization."
    },
    {
      "commit": "6c1b361daf3e36294363812ad378d7e052da1878",
      "tree": "4c8ad15f1d4d0e9d486f7a90a57f6ca5101a6145",
      "parents": [
        "15037af24371a0656d70ee60f2d002a3249a8559"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Thu Nov 12 14:13:29 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 14:13:29 2026 +0100"
      },
      "message": "NoSQL: Reuse untouched index containers during finalization (#6044)\n\nCatalog commits finalize indexes even when their contents have not changed, repeating embedded-index rebuilding and serialization unnecessarily.\n\nReturn the original container when the index is unmodified and no spill is required.\nPreserve finalization rules and existing empty-index behavior.\nAdd regression coverage for container reuse, reads, no-op removals, tombstones, and required spills."
    },
    {
      "commit": "15037af24371a0656d70ee60f2d002a3249a8559",
      "tree": "fae45809cb4c6c2997c08bdb501940598608abc8",
      "parents": [
        "385db3d006643cf0f87f46afbc34b26c8be7779b"
      ],
      "author": {
        "name": "Robert Stupp",
        "email": "snazy@snazy.de",
        "time": "Thu Nov 12 14:13:01 2026 +0100"
      },
      "committer": {
        "name": "GitHub",
        "email": "noreply@github.com",
        "time": "Thu Nov 12 14:13:01 2026 +0100"
      },
      "message": "NoSQL: Avoid mapping entities twice during non-renaming updates (#6043)\n\nChanged entities were mapped once for no-op detection and again to build the replacement, repeating property copying and type-specific mapping.\n\nReuse the mapped builder for the replacement while preserving timestamp, part-count, ID, and version handling. Add regression coverage for unchanged and changed updates."
    }
  ],
  "next": "385db3d006643cf0f87f46afbc34b26c8be7779b"
}
