Alex Chi Z. 4c4cb80186 fix(pageserver): fix gc-compaction racing with legacy gc (#10052) ...

## Problem

close https://github.com/neondatabase/neon/issues/10049, close
https://github.com/neondatabase/neon/issues/10030, close
https://github.com/neondatabase/neon/issues/8861

part of https://github.com/neondatabase/neon/issues/9114

The legacy gc process calls `get_latest_gc_cutoff`, which uses a Rcu
different than the gc_info struct. In the gc_compaction_smoke test case,
the "latest" cutoff could be lower than the gc_info struct, causing
gc-compaction to collect data that could be accessed by
`latest_gc_cutoff`. Technically speaking, there's nothing wrong with
gc-compaction using gc_info without considering latest_gc_cutoff,
because gc_info is the source of truth. But anyways, let's fix it.

## Summary of changes

* gc-compaction uses `latest_gc_cutoff` instead of gc_info to determine
the gc horizon.
* if a gc-compaction is scheduled via tenant compaction iteration, it
will take the gc_block lock to avoid racing with functionalities like
detach ancestor (if it's triggered via manual compaction API without
scheduling, then it won't take the lock)

---------

Signed-off-by: Alex Chi Z <chi@neon.tech>
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>

2024-12-09 20:06:06 +00:00

..

data/extension_test/5670669815

[compute/postgres] feature: PostgreSQL 17 (#8573)

2024-09-12 23:18:41 +01:00

test_ancestor_branch.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_attach_tenant_config.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_auth_broker.py

[auth_broker]: regress test (#9541)

2024-10-29 11:39:09 +00:00

test_auth.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_backpressure.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_bad_connection.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_basebackup_error.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_branch_and_gc.py

test_runner: skip more tests using decorator instead of pytest.skip (#9704)

2024-11-11 18:07:01 +00:00

test_branch_behind.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_branching.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_broken_timeline.py

pageserver: fix spurious error logs in timeline lifecycle (#9589)

2024-10-31 14:44:59 +00:00

test_build_info_metric.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_change_pageserver.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_clog_truncate.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_close_fds.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_combocid.py

test_runner: use LFC by default (#8613)

2024-11-25 09:01:05 +00:00

test_compaction.py

fix(pageserver): fix gc-compaction racing with legacy gc (#10052)

2024-12-09 20:06:06 +00:00

test_compatibility.py

Python 3.11 (#9515)

2024-11-21 16:25:31 +00:00

test_compute_catalog.py

fix(compute_ctl): Allow usage of DB names with whitespaces (#9919)

2024-11-28 21:38:30 +00:00

test_compute_locales.py

Switch compute-related locales to C.UTF-8 by default

2024-11-08 12:19:18 -06:00

test_compute_metrics.py

Bump sql_exporter to 0.16.0 (#10041)

2024-12-06 17:43:55 +00:00

test_config.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_crafted_wal_end.py

safekeeper: use protobuf for sending compressed records to pageserver (#9821)

2024-11-27 12:12:21 +00:00

test_createdropdb.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_createuser.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_ddl_forwarding.py

build(deps): bump mypy from 1.3.0 to 1.13.0 (#9670)

2024-11-22 14:31:36 +00:00

test_disk_usage_eviction.py

tests & benchmarks: unify the way we customize the default tenant config (#9992)

2024-12-03 22:07:03 +00:00

test_download_extensions.py

test_runner: skip more tests using decorator instead of pytest.skip (#9704)

2024-11-11 18:07:01 +00:00

test_endpoint_crash.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_explain_with_lfc_stats.py

test_runner: use LFC by default (#8613)

2024-11-25 09:01:05 +00:00

test_extensions.py

compute_ctl: Add endpoint that allows extensions to be installed (#9344)

2024-10-18 15:07:36 +03:00

test_fsm_truncate.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_fullbackup.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_gc_aggressive.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_gin_redo.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_hot_standby.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_import_pgdata.py

fast imports: initial Importer and Storage changes (#9218)

2024-11-22 22:47:06 +00:00

test_import.py

pageserver: fix spurious error logs in timeline lifecycle (#9589)

2024-10-31 14:44:59 +00:00

test_ingestion_layer_size.py

Python 3.11 (#9515)

2024-11-21 16:25:31 +00:00

test_installed_extensions.py

Rename 'installed_extensions' metric to 'compute_installed_extensions' (#9759)

2024-11-22 19:27:04 +00:00

test_large_schema.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_layer_bloating.py

test_runner: skip more tests using decorator instead of pytest.skip (#9704)

2024-11-11 18:07:01 +00:00

test_layer_eviction.py

test_runner: skip more tests using decorator instead of pytest.skip (#9704)

2024-11-11 18:07:01 +00:00

test_layer_writers_fail.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_layers_from_future.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_lfc_resize.py

test_runner: use LFC by default (#8613)

2024-11-25 09:01:05 +00:00

test_lfc_working_set_approximation.py

test_runner: use LFC by default (#8613)

2024-11-25 09:01:05 +00:00

test_local_file_cache.py

test_runner: use LFC by default (#8613)

2024-11-25 09:01:05 +00:00

test_logging.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_logical_replication.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_lsn_mapping.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_migrations.py

Grant access to pg_show_replication_origin_status for neon_superuser

2024-09-16 16:38:55 +01:00

test_multixact.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_neon_cli.py

test_runner: skip more tests using decorator instead of pytest.skip (#9704)

2024-11-11 18:07:01 +00:00

test_neon_extension.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_neon_local_cli.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_neon_superuser.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_next_xid.py

Update ruff to much newer version (#9433)

2024-10-18 12:42:41 +02:00

test_normal_work.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_oid_overflow.py

test_runner: use LFC by default (#8613)

2024-11-25 09:01:05 +00:00

test_old_request_lsn.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_ondemand_download.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_ondemand_slru_download.py

Python 3.11 (#9515)

2024-11-21 16:25:31 +00:00

test_ondemand_wal_download.py

Register custom xlog reader callbacks for on-demand WAL download in StartupDecodingContext (#9007)

2024-11-19 22:29:57 +02:00

test_pageserver_api.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_pageserver_catchup.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_pageserver_crash_consistency.py

fix(test_regress): always use storcon virtual pageserver API to set tenant config (#9622)

2024-11-04 17:42:08 +01:00

test_pageserver_generations.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_pageserver_getpage_throttle.py

tests & benchmarks: unify the way we customize the default tenant config (#9992)

2024-12-03 22:07:03 +00:00

test_pageserver_layer_rolling.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_pageserver_metric_collection.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_pageserver_reconnect.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_pageserver_restart.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_pageserver_restarts_under_workload.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_pageserver_secondary.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_parallel_copy.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_pg_query_cancellation.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_pg_regress.py

Python 3.11 (#9515)

2024-11-21 16:25:31 +00:00

test_pg_waldump.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_physical_and_logical_replicaiton.py

Do not wallog AUX files at replica (#9457)

2024-11-13 08:50:01 +02:00

test_physical_replication.py

Add GUC controlling whether to pause recovery if some critical GUCs at replica have smaller value than on primary (#9057)

2024-12-01 12:23:10 +00:00

test_pitr_gc.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_postgres_version.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_prefetch_buffer_resize.py

Fix issues with prefetch ring buffer resize (#9847)

2024-12-01 15:47:28 +00:00

test_proxy_allowed_ips.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_proxy_metric_collection.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_proxy_websockets.py

Improve type safety according to pyright

2024-11-08 14:43:15 -06:00

test_proxy.py

feat(proxy): add option to forward startup params (#9979)

2024-12-04 12:58:35 +00:00

test_read_validation.py

test_runner: use LFC by default (#8613)

2024-11-25 09:01:05 +00:00

test_readonly_node.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_recovery.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_remote_storage.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_replica_start.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_role_grants.py

compute_ctl: Add endpoint that allows setting role grants (#9395)

2024-10-18 11:25:45 +01:00

test_s3_restore.py

Python 3.11 (#9515)

2024-11-21 16:25:31 +00:00

test_setup.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_sharding.py

storcon: use proper schedule context during node delete (#9958)

2024-12-03 08:59:38 +00:00

test_sni_router.py

build(deps): bump mypy from 1.3.0 to 1.13.0 (#9670)

2024-11-22 14:31:36 +00:00

test_storage_controller.py

storcon: automatically clear Pause/Stop scheduling policies to enable detaches (#10011)

2024-12-07 13:05:09 +00:00

test_storage_scrubber.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_subscriber_restart.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_subxacts.py

safekeeper: use protobuf for sending compressed records to pageserver (#9821)

2024-11-27 12:12:21 +00:00

test_tenant_conf.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_tenant_delete.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_tenant_detach.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_tenant_relocation.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_tenant_size.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_tenant_tasks.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_tenants_with_remote_storage.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_tenants.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_threshold_based_eviction.py

Improve type safety according to pyright

2024-11-08 14:43:15 -06:00

test_timeline_archive.py

Support tenant manifests in the scrubber (#9942)

2024-12-03 20:39:10 +00:00

test_timeline_delete.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_timeline_detach_ancestor.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_timeline_gc_blocking.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_timeline_size.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_truncate.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_twophase.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_unlogged.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00

test_unstable_extensions.py

Create the notion of unstable extensions

2024-10-28 17:47:15 -05:00

test_vm_bits.py

test_runner: use LFC by default (#8613)

2024-11-25 09:01:05 +00:00

test_vm_truncate.py

Correctly truncate VM (#9342)

2024-11-14 17:19:13 +02:00

test_wal_acceptor_async.py

safekeeper: use protobuf for sending compressed records to pageserver (#9821)

2024-11-27 12:12:21 +00:00

test_wal_acceptor.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_wal_receiver.py

test_runner: improve wait_until (#9936)

2024-12-02 10:26:15 +00:00

test_wal_restore.py

Switch compute-related locales to C.UTF-8 by default

2024-11-08 12:19:18 -06:00

test_walredo_not_left_behind_on_detach.py

Enable all pyupgrade checks in ruff

2024-10-08 14:32:26 -05:00