issues

Messages by Date

2026/10/30 Re: [PR] Core: Add MutableBitmap.merge [iceberg] rdblue (via GitHub) via issues
2026/10/30 Re: [PR] Core: Add MutableBitmap.merge [iceberg] rdblue (via GitHub) via issues
2026/10/30 Re: [PR] Core: Add MutableBitmap.merge [iceberg] rdblue (via GitHub) via issues
2026/10/30 Re: [PR] Core: Add MutableBitmap.merge [iceberg] rdblue (via GitHub) via issues
2026/10/30 Re: [PR] Core: Add MutableBitmap.merge [iceberg] rdblue (via GitHub) via issues
2026/10/30 Re: [I] `dynamic_partition_overwrite` by `identity(s.x)` deletes rows by top-level `x` [iceberg-python] via GitHub
2026/10/30 Re: [PR] Core: Add MutableBitmap.merge [iceberg] rdblue (via GitHub) via issues
2026/10/30 Re: [PR] Spark: Compare filtered scan results by their unique record IDs so tests do not depend on Spark task or row ordering [iceberg] via GitHub
2026/10/30 Re: [I] Spark: WAP snapshot isolation broken for format-version 3 tables when deletion vectors are used (spark.wap.id) [iceberg] via GitHub
2026/10/30 Re: [PR] Core: Add MutableBitmap.merge [iceberg] rdblue (via GitHub) via issues
2026/10/30 Re: [PR] API: [1/N] Add ReadRestrictions Actions [iceberg] huaxingao (via GitHub) via issues
2026/10/30 Re: [PR] Core: Add MutableBitmap.merge [iceberg] rdblue (via GitHub) via issues
2026/10/30 Re: [I] Parquet: Page Index pruning in the custom Parquet reader [iceberg] yangshangqing95 (via GitHub) via issues
2026/10/30 Re: [PR] Parquet: Preserve non-contiguous row positions in vectorized reads [iceberg] yangshangqing95 (via GitHub) via issues
2026/10/30 [PR] Spark 4.0, 4.1: Backport: Forward parseRoutineParam and parseQuery to the delegate parser (#18702) [iceberg] via GitHub
2026/10/30 Re: [PR] Flink: Distribute rows by custom fields to DynamicWriter in DynamicSink [iceberg] via GitHub
2026/10/30 [PR] dev: add ty as an experimental type checker [iceberg-python] via GitHub
2026/10/30 Re: [PR] fix: handle more types of timestamp conversion [iceberg-rust] mbutrovich (via GitHub) via issues
2026/10/30 Re: [PR] Spark 4.0, 4.1: Backport: Forward parseRoutineParam and parseQuery to the delegate parser (#18702) [iceberg] huaxingao (via GitHub) via issues
2026/10/30 Re: [PR] fix: handle more types of timestamp conversion [iceberg-rust] mbutrovich (via GitHub) via issues
2026/10/30 Re: [PR] Spark 4.0, 4.1: Backport: Forward parseRoutineParam and parseQuery to the delegate parser (#18702) [iceberg] via GitHub
2026/10/30 [PR] fix(arrow): decode data files on the CPU runtime [iceberg-rust] via GitHub
2026/10/30 Re: [PR] Core: Extract commit-conflict validations into SnapshotValidator [iceberg] anoopj (via GitHub) via issues
2026/10/30 Re: [PR] Flink: Resolve the record serializer without loading the table [iceberg] via GitHub
2026/10/30 Re: [PR] Spark 3.5, 4.0, 4.1, 4.2: Use scan schema for CBO column statistics [iceberg] via GitHub
2026/10/30 Re: [PR] API, Core: Implement v4 file-level tight_stats [iceberg] amogh-jahagirdar (via GitHub) via issues
2026/10/30 Re: [PR] feat(transaction): Add MergingSnapshotProducer and ManifestFilter [iceberg-rust] CTTY (via GitHub) via issues
2026/10/30 Re: [PR] feat(transaction): Add MergingSnapshotProducer and ManifestFilter [iceberg-rust] CTTY (via GitHub) via issues
2026/10/30 Re: [PR] feat(table): plan delete-file changes in incremental changelog scans [iceberg-go] via GitHub
2026/10/30 Re: [PR] Core: Extract commit-conflict validations into SnapshotValidator [iceberg] anoopj (via GitHub) via issues
2026/10/30 [PR] Fix snapshot history after expiring snapshots [iceberg-python] AMBRA7592 (via GitHub) via issues
2026/10/30 Re: [PR] API, Core: Implement v4 file-level tight_stats [iceberg] amogh-jahagirdar (via GitHub) via issues
2026/10/30 [I] CPU work done on IO runtime [iceberg-rust] alexanderbianchi (via GitHub) via issues
2026/10/30 [PR] Build: Bump joserfc from 1.7.2 to 1.7.6 [iceberg-python] dependabot[bot] (via GitHub) via issues
2026/10/30 Re: [PR] Core: Extract commit-conflict validations into SnapshotValidator [iceberg] anoopj (via GitHub) via issues
2026/10/30 Re: [PR] Core: Extract commit-conflict validations into SnapshotValidator [iceberg] anoopj (via GitHub) via issues
2026/10/30 Re: [PR] Core: Extract commit-conflict validations into SnapshotValidator [iceberg] anoopj (via GitHub) via issues
2026/10/30 [PR] Fix duplicate sort order id when adding an existing sort order [iceberg-python] via GitHub
2026/10/30 Re: [PR] Core: Introduce content_stats [iceberg] rdblue (via GitHub) via issues
2026/10/30 Re: [PR] Flink: Resolve the record serializer without loading the table [iceberg] via GitHub
2026/10/30 Re: [PR] Spark 3.5, 4.0, 4.1, 4.2: Use scan schema for CBO column statistics [iceberg] via GitHub
2026/10/30 [PR] Fix row filter literals narrowed to old file types after type promotion [iceberg-python] via GitHub
2026/10/30 Re: [PR] Spec: Add V4 column updates to the spec [iceberg] via GitHub
2026/10/30 Re: [PR] Data, Spark 4.2: Add format model TCK coverage for vectorized reads [iceberg] via GitHub
2026/10/30 [PR] dev: move every check into one noxfile [iceberg-python] via GitHub
2026/10/30 Re: [PR] Spark: Support aggregate pushdown for identity partition column GROUP BY [iceberg] via GitHub
2026/10/30 Re: [PR] fix: make In and NotIn NULL for a NULL value in the row filter [iceberg-go] via GitHub
2026/10/30 Re: [PR] feat: cache immutable metadata content in FileIO [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] fix(spec): format dropped-source integer partition values [iceberg-rust] via GitHub
2026/10/30 [PR] AWS: Fix DynamoDbCatalog listNamespaces to return only child namespaces [iceberg] via GitHub
2026/10/30 [PR] feat(transaction): add MergingSnapshotProducer and RewriteFilesAction (continues #3046) [iceberg-rust] via GitHub
2026/10/30 Re: [I] rest: decide whether a partial vended credential replaces or layers onto the base config [iceberg-go] via GitHub
2026/10/30 Re: [PR] Kafka Connect: Make KafkaMetadataTransform configuration per-instance [iceberg] via GitHub
2026/10/30 Re: [PR] feat(inspect): support selecting columns in metadata table scans [iceberg-cpp] via GitHub
2026/10/30 [PR] feat(encryption): add EncryptionManager and StandardEncryptionManager [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] Spec: Add collations for string types [iceberg] via GitHub
2026/10/30 [PR] fix(arrow): decode delete files on the CPU runtime [iceberg-rust] via GitHub
2026/10/30 Re: [PR] Fix inspect.entries() for data files without column_sizes [iceberg-python] via GitHub
2026/10/30 Re: [PR] fix: make In and NotIn NULL for a NULL value in the row filter [iceberg-go] via GitHub
2026/10/30 Re: [PR] Flink: Allow configuring a branch for rewrite data files action [iceberg] via GitHub
2026/10/30 Re: [PR] Flink: Allow configuring a branch for rewrite data files action [iceberg] via GitHub
2026/10/30 Re: [PR] fix(table): make NotIn skip NULL rows in scans [iceberg-go] via GitHub
2026/10/30 [PR] Flink: ConvertEqualityDeletes: Add support for incremental re-indexing [iceberg] via GitHub
2026/10/30 Re: [PR] feat(RTAS): add TableMetadataBuilder::build_replacement [iceberg-rust] via GitHub
2026/10/30 Re: [PR] feat(view): add view metadata types and validation [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] feat(data): enforce residual filters in the task reader [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] perf(rest): avoid repeated schema-history clones in partition decoding [iceberg-go] via GitHub
2026/10/30 Re: [PR] feat(scan): plan changelog tasks from deletion vectors [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] fix: handle more types of timestamp conversion [iceberg-rust] via GitHub
2026/10/30 [PR] Support deleting unreferenced files after snapshot expiration [iceberg-python] via GitHub
2026/10/30 Re: [PR] perf(internal): avoid quadratic packing with large lookback [iceberg-go] via GitHub
2026/10/30 Re: [PR] perf(table): bound partition metadata record batches [iceberg-go] via GitHub
2026/10/30 Re: [PR] fix(table): clean up expired files from planning metadata [iceberg-go] via GitHub
2026/10/30 Re: [I] Parquet: Page Index pruning in the custom Parquet reader [iceberg] via GitHub
2026/10/30 Re: [PR] feat(avro,parquet): read the full file schema when no projection is set [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] Core: Skip statistics removal changes when suppressing snapshots [iceberg] via GitHub
2026/10/30 Re: [I] `snapshot_as_of_timestamp` returns an older snapshot after `expire_snapshots` [iceberg-python] via GitHub
2026/10/30 Re: [I] `incremental_append_scan` from an expired start fails after `expire_snapshots` [iceberg-python] via GitHub
2026/10/30 [PR] Assert the last assigned field id when update_schema adds a schema [iceberg-python] via GitHub
2026/10/30 [PR] Keep the parent snapshot id when removing snapshots [iceberg-python] via GitHub
2026/10/30 Re: [PR] feat(inspect): support selecting columns in metadata table scans [iceberg-cpp] via GitHub
2026/10/30 [PR] perf(table): skip sorting already ordered compaction batches [iceberg-go] via GitHub
2026/10/30 [PR] refactorstore binary and fixed defaults as hex strings in schemas [iceberg-go] via GitHub
2026/10/30 Re: [I] Support configuring datafusion catalog in sqllogictest framework [iceberg-rust] via GitHub
2026/10/30 Re: [PR] Arrow, Spark: Skip unprojected shredded variants in vectorized Parquet reads [iceberg] via GitHub
2026/10/30 [PR] feat(catalog): support isolated FileIO registries [iceberg-go] via GitHub
2026/10/30 Re: [PR] feat(avro,parquet): read the full file schema when no projection is set [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] fix: handle more types of timestamp conversion [iceberg-rust] via GitHub
2026/10/30 Re: [I] A stale `add_column` reuses a dropped column's field id and schema id [iceberg-python] via GitHub
2026/10/30 Re: [PR] Spark 4.2: Use snapshot schema for CBO column statistics [iceberg] via GitHub
2026/10/30 Re: [PR] fix: handle more types of timestamp conversion [iceberg-rust] via GitHub
2026/10/30 Re: [PR] fix: handle more types of timestamp conversion [iceberg-rust] via GitHub
2026/10/30 Re: [PR] fix: handle more types of timestamp conversion [iceberg-rust] via GitHub
2026/10/30 Re: [PR] Spec: Add V4 column updates to the spec [iceberg] via GitHub
2026/10/30 [PR] GCP: Select refreshed GCS credential by storage prefix [iceberg] via GitHub
2026/10/30 Re: [I] feat: implement ReplacePartitions (two-PR plan) [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] feat(storage)(test): Add shared integration test suite for iceberg-storage (Supersedes #2253) [iceberg-rust] via GitHub
2026/10/30 Re: [PR] fix: handle more types of timestamp conversion [iceberg-rust] via GitHub
2026/10/30 Re: [I] Spark: WAP snapshot isolation broken for format-version 3 tables when deletion vectors are used (spark.wap.id) [iceberg] via GitHub
2026/10/30 Re: [PR] Flink: ConvertEqualityDeletes: Fix index consistency after a failure or restore [iceberg] via GitHub
2026/10/30 Re: [PR] Flink: ConvertEqualityDeletes: Fix index consistency after a failure or restore [iceberg] via GitHub
2026/10/30 [PR] Use the full column name in the dynamic_partition_overwrite delete filter [iceberg-python] via GitHub
2026/10/30 Re: [PR] Fix inspect.entries() for data files without column_sizes [iceberg-python] via GitHub
2026/10/30 Re: [PR] refactor(puffin): centralize writer state checks [iceberg-go] via GitHub
2026/10/30 [I] After promoting a `float` column to `double`, `>` and `!=` filters drop rows from older files [iceberg-python] via GitHub
2026/10/30 [I] Spark: WAP snapshot isolation broken for format-version 3 tables when deletion vectors are used (spark.wap.id) [iceberg] via GitHub
2026/10/30 Re: [PR] fix(arrow): make full batch validation configurable [iceberg-cpp] via GitHub
2026/10/30 Re: [I] Remove deleted data files with expire_snapshots [iceberg-python] via GitHub
2026/10/30 [PR] Core: Stop local scan planning early when minRowsRequested is satisfied [iceberg] via GitHub
2026/10/30 Re: [PR] feat(transaction): add overwrite action (continues #2185) [iceberg-rust] via GitHub
2026/10/30 Re: [PR] feat(transaction): add overwrite action (continues #2185) [iceberg-rust] via GitHub
2026/10/30 Re: [PR] feat(transaction): add overwrite action (continues #2185) [iceberg-rust] via GitHub
2026/10/30 [I] Dropped decimal partition source renders its unscaled value [iceberg-rust] via GitHub
2026/10/30 Re: [PR] feat(transaction): add MergingSnapshotProducer and RewriteFilesAction [iceberg-rust] via GitHub
2026/10/30 Re: [PR] fix(spec): format dropped-source integer partition values [iceberg-rust] via GitHub
2026/10/30 Re: [I] Umbrella: API and dependency cleanup before the v1.0.0 freeze [iceberg-go] via GitHub
2026/10/30 Re: [PR] feat(read): read identity partition columns from file scan tasks [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] fix(sql): redact passwords in connection URI errors [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] fix(arrow): make full batch validation configurable [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] feat(rest): support isolated IO registries [iceberg-go] via GitHub
2026/10/30 Re: [PR] fix(table): check partial-progress batches against the planning snapshot [iceberg-go] via GitHub
2026/10/30 Re: [PR] Flink: ConvertEqualityDeletes: Fix index consistency after a failure or restore [iceberg] via GitHub
2026/10/30 Re: [PR] feat(storage): Implement ObjectStoreStorage::S3 (Supersedes #2257) [iceberg-rust] via GitHub
2026/10/30 Re: [PR] feat(transaction): Add MergingSnapshotProducer and ManifestFilter [iceberg-rust] via GitHub
2026/10/30 Re: [I] Spark: WAP snapshot isolation broken for format-version 3 tables when deletion vectors are used (spark.wap.id) [iceberg] via GitHub
2026/10/30 Re: [PR] fix(arrow): make full batch validation configurable [iceberg-cpp] via GitHub
2026/10/30 [I] `expire_snapshots` on a stale table deletes a branch moved onto the expired snapshot [iceberg-python] via GitHub
2026/10/30 [I] A stale `update_sort_order` adds a second sort order with an existing id, which Java cannot load [iceberg-python] via GitHub
2026/10/30 Re: [PR] feat: apply write defaults in DataWriter [iceberg-cpp] via GitHub
2026/10/30 Re: [PR] feat(storage): Implement ObjectStoreStorage::S3 (Supersedes #2257) [iceberg-rust] via GitHub
2026/10/30 Re: [PR] feat(storage): Implement ObjectStoreStorage::S3 (Supersedes #2257) [iceberg-rust] via GitHub
2026/10/30 [PR] Docs: Fix type mismatch in JDBC catalog Java API example [iceberg] via GitHub
2026/10/30 [PR] Docs: Fix result name in Delta Lake migration output table [iceberg] via GitHub
2026/10/30 Re: [I] Docs: Delta Lake migration Output table names a result that does not exist [iceberg] via GitHub
2026/10/30 Re: [I] Aliyun: RRSA environment auto-detection overrides explicit static credentials [iceberg] via GitHub
2026/10/30 [PR] AWS: Apply S3 service configurations to async S3 clients [iceberg] via GitHub
2026/10/30 Re: [I] AWS: DefaultAwsClientFactory s3 / s3Async serviceConfiguration are not equivalent [iceberg] via GitHub
2026/10/30 Re: [PR] Spark: Fix ClassLoader leak in Spark caused by AuthSessionCache and ensure resource cleanup [iceberg] via GitHub
2026/10/30 [PR] Fix inspect.entries() for data files without column_sizes [iceberg-python] via GitHub
2026/10/30 Re: [PR] Flink: Backport: Close the Sink V2 aggregator's table loader (#18496) [iceberg] via GitHub
2026/10/30 Re: [PR] ci: resolve Arrow and Parquet from conda-forge on Unix legs [iceberg-cpp] via GitHub
2026/10/30 [I] A stale `update_spec` adds a partition field named like a column another writer added, which Java cannot load [iceberg-python] via GitHub
2026/10/30 Re: [I] Support geo types [iceberg-rust] via GitHub
2026/10/29 Re: [PR] Flink: Backport: Forward watermarks in EqualityConvertCommitter with no active cycle (#18439) [iceberg] via GitHub
2026/10/29 Re: [PR] Data, Spark 4.2: Add format model TCK coverage for vectorized reads [iceberg] via GitHub
2026/10/29 Re: [PR] feat: add KeyManagementClient, KmsRegistry and InMemoryKms [iceberg-cpp] via GitHub
2026/10/29 Re: [PR] feat: add KeyManagementClient, KmsRegistry and InMemoryKms [iceberg-cpp] via GitHub
2026/10/29 [PR] fix: keep NaN rows when RewriteNotExpr pushes a NOT down [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(catalog): support write.metadata.delete-after-commit.enabled [iceberg-rust] via GitHub
2026/10/29 [I] `dynamic_partition_overwrite` by `identity(s.x)` deletes rows by top-level `x` [iceberg-python] via GitHub
2026/10/29 [I] `count()` and `plan_files()` return nothing for an expired snapshot id [iceberg-python] via GitHub
2026/10/29 Re: [PR] BigQuery: Align metastore default read timeout with the SDK [iceberg] via GitHub
2026/10/29 Re: [PR] Data, Spark 4.2: Add format model TCK coverage for vectorized reads [iceberg] via GitHub
2026/10/29 Re: [PR] Data, Spark 4.2: Add format model TCK coverage for vectorized reads [iceberg] via GitHub
2026/10/29 Re: [PR] Data, Spark 4.2: Add format model TCK coverage for vectorized reads [iceberg] via GitHub
2026/10/29 Re: [PR] Data, Spark 4.2: Add format model TCK coverage for vectorized reads [iceberg] via GitHub
2026/10/29 Re: [PR] Data, Spark 4.2: Add format model TCK coverage for vectorized reads [iceberg] via GitHub
2026/10/29 [I] Appending to a branch that does not exist creates it without main's rows [iceberg-python] via GitHub
2026/10/29 Re: [PR] Data, Spark 4.2: Add format model TCK coverage for vectorized reads [iceberg] via GitHub
2026/10/29 Re: [PR] Data, Spark 4.2: Add format model TCK coverage for vectorized reads [iceberg] via GitHub
2026/10/29 Re: [I] A retried append records its file under a concurrently added partition spec [iceberg-python] via GitHub
2026/10/29 [I] A retried `upsert` inserts a key that a concurrent writer already inserted [iceberg-python] via GitHub
2026/10/29 [I] `incremental_append_scan` from an expired start fails after `expire_snapshots` [iceberg-python] via GitHub
2026/10/29 [I] `dynamic_partition_overwrite` after an in-transaction `update_spec` loses rows [iceberg-python] via GitHub
2026/10/29 [PR] Spark: Report effective split target size as a driver metric [iceberg] via GitHub
2026/10/29 [PR] fix: make In and NotIn NULL for a NULL value in the row filter [iceberg-go] via GitHub
2026/10/29 [I] `add_files` puts a file holding a value and NULLs in that value's partition [iceberg-python] via GitHub
2026/10/29 [I] Spark: Expose the effective split target size as a driver metric [iceberg] via GitHub
2026/10/29 Re: [PR] Core: Add MutableBitmap.merge [iceberg] via GitHub
2026/10/29 [I] A stale `add_column` reuses a dropped column's field id and schema id [iceberg-python] via GitHub
2026/10/29 Re: [I] `create_branch` and `create_tag` replace an existing ref of the same name [iceberg-python] via GitHub
2026/10/29 [I] A long transaction committed past a concurrent tag writes metadata Java rejects [iceberg-python] via GitHub
2026/10/29 Re: [PR] fix(transform): match Java truncate wrapping arithmetic [iceberg-rust] via GitHub
2026/10/29 [I] A retried append records its file under a concurrently added partition spec [iceberg-python] via GitHub
2026/10/29 [I] Changelog planning still scans every live data manifest when a snapshot adds deletes [iceberg-go] via GitHub
2026/10/29 [I] `create_branch` and `create_tag` replace an existing ref of the same name [iceberg-python] via GitHub
2026/10/29 Re: [PR] feat(table): read rows removed by incremental changelog deletes [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(table): plan delete-file changes in incremental changelog scans [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(table): read file-level incremental changelog rows [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(table): plan delete-file changes in incremental changelog scans [iceberg-go] via GitHub
2026/10/29 Re: [PR] ci: resolve Arrow and Parquet from conda-forge on Unix legs [iceberg-cpp] via GitHub
2026/10/29 Re: [PR] feat(table): read file-level incremental changelog rows [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(table): read rows removed by incremental changelog deletes [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(table): read rows removed by incremental changelog deletes [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(table): plan delete-file changes in incremental changelog scans [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(table): read file-level incremental changelog rows [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(table): read file-level incremental changelog rows [iceberg-go] via GitHub
2026/10/29 Re: [PR] feat(table): plan delete-file changes in incremental changelog scans [iceberg-go] via GitHub
2026/10/29 [I] Read changelog files in parallel while keeping task order [iceberg-go] via GitHub
2026/10/29 Re: [PR] Core: Add MutableBitmap.merge [iceberg] via GitHub
2026/10/29 Re: [PR] fix(upsert): bound file match filter to avoid segfault [iceberg-python] via GitHub
2026/10/29 Re: [PR] Core: Extract commit-conflict validations into SnapshotValidator [iceberg] via GitHub
2026/10/29 Re: [PR] Spark: Support aggregate pushdown for identity partition column GROUP BY [iceberg] via GitHub
2026/10/29 Re: [PR] test(storage-opendal): split backend tests into per-backend modules [iceberg-rust] via GitHub
2026/10/29 Re: [PR] Flink: ConvertEqualityDeletes: Fix index consistency after a failure or restore [iceberg] via GitHub
2026/10/29 Re: [PR] Core, REST: Avoid loading all snapshots when committing in refs mode [iceberg] via GitHub
2026/10/29 [PR] fix(arrow): keep the parquet error as the source of a failed delete file read [iceberg-rust] via GitHub
2026/10/29 Re: [PR] feat(io): route FileIO operations through the configured runtime [iceberg-rust] via GitHub
2026/10/29 Re: [PR] Core: Add util to derive source column bounds from monotonic partition transforms [iceberg] via GitHub
2026/10/29 Re: [PR] Spark 4.2: Best-effort ExpireSnapshots when metadata reachable only from expired snapshots is missing [iceberg] via GitHub