Package: release.debian.org Severity: normal Tags: trixie User: [email protected] Usertags: pu
Hi, I would like to propose the 4.0.7 stable point release of patroni (which has been in testing for three weeks) to trixie, upgrading from the current 4.0.6 version. Debdiff is attached, I have reviewed the upstream changes. Michael
diff -Nru patroni-4.0.6/debian/changelog patroni-4.0.7/debian/changelog --- patroni-4.0.6/debian/changelog 2025-06-08 08:52:19.000000000 +0200 +++ patroni-4.0.7/debian/changelog 2025-10-20 14:55:24.000000000 +0200 @@ -1,3 +1,35 @@ +patroni (4.0.7-3+deb13u1) trixie; urgency=medium + + * Upload to stable. + + -- Michael Banck <[email protected]> Mon, 20 Oct 2025 14:55:24 +0200 + +patroni (4.0.7-3) unstable; urgency=medium + + * debian/tests/acceptance: Really make the changes from the last revision + this time. + + -- Michael Banck <[email protected]> Thu, 25 Sep 2025 15:01:53 +0200 + +patroni (4.0.7-2) unstable; urgency=medium + + * debian/tests/acceptance: Further changes to stopping etcd. If the init + script exists, use it. Otherwise, if systemd is available, use that. If + neither are available, do not try to stop etcd. + + -- Michael Banck <[email protected]> Thu, 25 Sep 2025 14:20:40 +0200 + +patroni (4.0.7-1) unstable; urgency=medium + + * New upstream release. + * debian/patches/startup_scripts.patch: Refreshed. + * debian/patches/avoid_overwriting_configuration_during_boostrap.patch: + Likewise. + * debian/patches/replslot-cluster-type-attribute.patch: Likewise. + * debian/tests/acceptance: Only stop etcd if init script exists. + + -- Michael Banck <[email protected]> Wed, 24 Sep 2025 16:58:33 +0200 + patroni (4.0.6-1) unstable; urgency=medium * New upstream release. diff -Nru patroni-4.0.6/debian/patches/avoid_overwriting_configuration_during_boostrap.patch patroni-4.0.7/debian/patches/avoid_overwriting_configuration_during_boostrap.patch --- patroni-4.0.6/debian/patches/avoid_overwriting_configuration_during_boostrap.patch 2025-03-15 11:57:39.000000000 +0100 +++ patroni-4.0.7/debian/patches/avoid_overwriting_configuration_during_boostrap.patch 2025-09-24 08:10:12.000000000 +0200 @@ -21,7 +21,7 @@ =================================================================== --- patroni.orig/patroni/postgresql/config.py +++ patroni/patroni/postgresql/config.py -@@ -510,7 +510,7 @@ class ConfigHandler(object): +@@ -513,7 +513,7 @@ class ConfigHandler(object): try: for f in self._configuration_to_save: config_file = os.path.join(self._config_dir, f) @@ -30,7 +30,7 @@ if os.path.isfile(config_file): shutil.copy(config_file, backup_file) self.set_file_permissions(backup_file) -@@ -523,7 +523,7 @@ class ConfigHandler(object): +@@ -526,7 +526,7 @@ class ConfigHandler(object): try: for f in self._configuration_to_save: config_file = os.path.join(self._config_dir, f) diff -Nru patroni-4.0.6/debian/patches/replslot-cluster-type-attribute.patch patroni-4.0.7/debian/patches/replslot-cluster-type-attribute.patch --- patroni-4.0.6/debian/patches/replslot-cluster-type-attribute.patch 2025-02-20 22:32:02.000000000 +0100 +++ patroni-4.0.7/debian/patches/replslot-cluster-type-attribute.patch 2025-10-20 13:59:09.000000000 +0200 @@ -10,11 +10,11 @@ patroni/dcs/__init__.py | 4 +++- 2 files changed, 5 insertions(+), 2 deletions(-) -diff --git a/docs/dynamic_configuration.rst b/docs/dynamic_configuration.rst -index 09413312d..908bc3f68 100644 ---- a/docs/dynamic_configuration.rst -+++ b/docs/dynamic_configuration.rst -@@ -61,9 +61,10 @@ In order to change the dynamic configuration you can use either :ref:`patronictl +Index: patroni/docs/dynamic_configuration.rst +=================================================================== +--- patroni.orig/docs/dynamic_configuration.rst ++++ patroni/docs/dynamic_configuration.rst +@@ -61,9 +61,10 @@ In order to change the dynamic configura - **my\_slot\_name**: the name of the permanent replication slot. If the permanent slot name matches with the name of the current node it will not be created on this node. If you add a permanent physical replication slot which name matches the name of a Patroni member, Patroni will ensure that the slot that was created is not removed even if the corresponding member becomes unresponsive, situation which would normally result in the slot's removal by Patroni. Although this can be useful in some situations, such as when you want replication slots used by members to persist during temporary failures or when importing existing members to a new Patroni cluster (see :ref:`Convert a Standalone to a Patroni Cluster <existing_data>` for details), caution should be exercised by the operator that these clashes in names are not persisted in the DCS, when the slot is no longer required, due to its effect on normal functioning of Patroni. @@ -26,11 +26,11 @@ - **ignore\_slots**: list of sets of replication slot properties for which Patroni should ignore matching slots. This configuration/feature/etc. is useful when some replication slots are managed outside of Patroni. Any subset of matching properties will cause a slot to be ignored. -diff --git a/patroni/dcs/__init__.py b/patroni/dcs/__init__.py -index 4a3516426..19d3c1aa1 100644 ---- a/patroni/dcs/__init__.py -+++ b/patroni/dcs/__init__.py -@@ -998,7 +998,9 @@ def __permanent_slots(self) -> Dict[str, Union[Dict[str, Any], Any]]: +Index: patroni/patroni/dcs/__init__.py +=================================================================== +--- patroni.orig/patroni/dcs/__init__.py ++++ patroni/patroni/dcs/__init__.py +@@ -1011,7 +1011,9 @@ class Cluster(NamedTuple('Cluster', @property def permanent_physical_slots(self) -> Dict[str, Any]: """Dictionary of permanent ``physical`` replication slots.""" diff -Nru patroni-4.0.6/debian/patches/startup_scripts.patch patroni-4.0.7/debian/patches/startup_scripts.patch --- patroni-4.0.6/debian/patches/startup_scripts.patch 2024-11-09 10:47:13.000000000 +0100 +++ patroni-4.0.7/debian/patches/startup_scripts.patch 2025-10-20 13:59:09.000000000 +0200 @@ -55,8 +55,8 @@ - # Pre-commands to start watchdog device # Uncomment if watchdog is part of your patroni setup - #ExecStartPre=-/usr/bin/sudo /sbin/modprobe softdog - #ExecStartPre=-/usr/bin/sudo /bin/chown postgres /dev/watchdog + #ExecStartPre=-+/sbin/modprobe softdog + #ExecStartPre=-+/bin/chown postgres /dev/watchdog # Start the patroni process -ExecStart=/bin/patroni /etc/patroni.yml diff -Nru patroni-4.0.6/debian/patroni.service patroni-4.0.7/debian/patroni.service --- patroni-4.0.6/debian/patroni.service 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/debian/patroni.service 2025-09-22 18:07:47.000000000 +0200 @@ -23,8 +23,8 @@ # Pre-commands to start watchdog device # Uncomment if watchdog is part of your patroni setup -#ExecStartPre=-/usr/bin/sudo /sbin/modprobe softdog -#ExecStartPre=-/usr/bin/sudo /bin/chown postgres /dev/watchdog +#ExecStartPre=-+/sbin/modprobe softdog +#ExecStartPre=-+/bin/chown postgres /dev/watchdog # Start the patroni process ExecStart=/bin/patroni /etc/patroni.yml diff -Nru patroni-4.0.6/debian/tests/acceptance patroni-4.0.7/debian/tests/acceptance --- patroni-4.0.6/debian/tests/acceptance 2025-02-20 22:32:02.000000000 +0100 +++ patroni-4.0.7/debian/tests/acceptance 2025-09-25 14:18:30.000000000 +0200 @@ -49,8 +49,16 @@ # ensure no etcd server is running. if [ "$DCS" = "etcd" -o "$DCS" = "etcd3" ] then - service etcd stop - service etcd status || true + if [ -x /etc/init.d/etcd ] + then + service etcd stop + service etcd status || true + else + if [ -d /run/systemd/system ] + then + systemctl stop etcd + fi + fi fi # make sure various directories are writable by the postgres user for dir in features/output data /tmp/pgpass_postgres-?; do diff -Nru patroni-4.0.6/docs/index.rst patroni-4.0.7/docs/index.rst --- patroni-4.0.6/docs/index.rst 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/docs/index.rst 2025-09-22 18:07:47.000000000 +0200 @@ -10,7 +10,7 @@ We call Patroni a "template" because it is far from being a one-size-fits-all or plug-and-play replication system. It will have its own caveats. Use wisely. There are many ways to run high availability with PostgreSQL; for a list, see the `PostgreSQL Documentation <https://wiki.postgresql.org/wiki/Replication,_Clustering,_and_Connection_Pooling>`__. -Currently supported PostgreSQL versions: 9.3 to 17. +Currently supported PostgreSQL versions: 9.3 to 18. **Note to Citus users**: Starting from 3.0 Patroni nicely integrates with the `Citus <https://github.com/citusdata/citus>`__ database extension to Postgres. Please check the :ref:`Citus support page <citus>` in the Patroni documentation for more info about how to use Patroni high availability together with a Citus distributed cluster. diff -Nru patroni-4.0.6/docs/README.rst patroni-4.0.7/docs/README.rst --- patroni-4.0.6/docs/README.rst 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/docs/README.rst 2025-09-22 18:07:47.000000000 +0200 @@ -34,6 +34,10 @@ there is no requirement on the minimal number of nodes. Running a cluster consisting of one primary and one standby is perfectly fine. You can add more standby nodes later. +**2-node clusters** (primary + standby) are common and provide automatic failover with high availability. Note that during failover, you'll temporarily have no redundancy until the failed node rejoins. + +**DCS requirements**: Your DCS (etcd, ZooKeeper, Consul) has to run with **3 or 5 nodes** for proper consensus and fault tolerance. A single DCS cluster can store information for hundreds or thousands of Patroni clusters using different namespace/scope combinations. + Running and Configuring ----------------------- diff -Nru patroni-4.0.6/docs/releases.rst patroni-4.0.7/docs/releases.rst --- patroni-4.0.6/docs/releases.rst 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/docs/releases.rst 2025-09-22 18:07:47.000000000 +0200 @@ -3,6 +3,56 @@ Release notes ============= +Version 4.0.7 +------------- + +Released 2025-09-22 + +**New features** + +- Add support for PostgreSQL 18 RC1 (Alexander Kukushkin) + + GUC's validator rules were extended. Patroni now properly handles the new background I/O worker. + + +**Bugfixes** + +- Fix potential issue around resolving localhost to IPv6 on Windows (András Váczi) + + When configuring ``listen_addresses`` in PostgreSQL, using ``0.0.0.0`` or ``127.0.0.1`` will restrict listening to IPv4 only, excluding IPv6. On typical Windows systems, however, ``localhost`` often resolves to the IPv6 address ``::1`` by default. To ensure compatibility, Patroni now configures PostgreSQL to listen on ``127.0.0.1``, instead of ``localhost``, on Windows systems. + +- Return global config only when ``/config`` key exists in DCS (Alexander Kukushkin) + + Patroni REST API was returning an empty configuration instead of raising an error if the ``/config`` key was missing in DCS. + +- Fix the issue of failsafe mode not being triggered in case of Etcd unavailability (Alexander Kukushkin) + + Patroni was not always properly handling ``etcd3`` exceptions, which resulted in failsafe mode not being triggered. + +- Fix signal handler reentrancy deadlock (Waynerv) + + Patroni running in a Docker container with ``PID=1`` in some special cases was experiencing deadlock after receiving ``SIGCHLD``. + +- Recreate (permanent) physical slot when it doesn't reserve WAL (Israel Barth Rubio) + + Permanent physical replication slots created outside of Patroni scope without reserving WALs were causing a ``replication slot cannot be advanced`` error. To avoid this, Patroni now recreates such slots. + +- Handle watch cancelation messages in ``etcd3`` properly (Alexander Kukushkin) + + When ``etcd3`` sends a cancelation message to the watch channel, it doesn't close the connection. This results in Patroni using stale data. Patroni now solves it by breaking a loop of reading chunked response and closing the connection on the Patroni side. + +- Handle case when ``HTTPConnection`` socket is wrapped with ``pyopenssl`` (Alexander Kukushkin) + + Patroni was not correctly using ``pyopenssl`` interfaces, enforced in ``python-etcd``. + + +**Documentation improvements** + +- Improve 2-node cluster guidance (Nikolay Samokhvalov) + + Clarify behaviour during failover and DCS requirements. + + Version 4.0.6 ------------- diff -Nru patroni-4.0.6/extras/startup-scripts/patroni.service patroni-4.0.7/extras/startup-scripts/patroni.service --- patroni-4.0.6/extras/startup-scripts/patroni.service 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/extras/startup-scripts/patroni.service 2025-09-22 18:07:47.000000000 +0200 @@ -23,8 +23,8 @@ # Pre-commands to start watchdog device # Uncomment if watchdog is part of your patroni setup -#ExecStartPre=-/usr/bin/sudo /sbin/modprobe softdog -#ExecStartPre=-/usr/bin/sudo /bin/chown postgres /dev/watchdog +#ExecStartPre=-+/sbin/modprobe softdog +#ExecStartPre=-+/bin/chown postgres /dev/watchdog # Start the patroni process ExecStart=/bin/patroni /etc/patroni.yml diff -Nru patroni-4.0.6/features/steps/basic_replication.py patroni-4.0.7/features/steps/basic_replication.py --- patroni-4.0.6/features/steps/basic_replication.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/features/steps/basic_replication.py 2025-09-22 18:07:47.000000000 +0200 @@ -102,7 +102,7 @@ assert False, "Error loading test data on {0}: {1}".format(pg_name, e) -@then('Table {table_name:w} is present on {pg_name:name} after {max_replication_delay:d} seconds') +@then('table {table_name:w} is present on {pg_name:name} after {max_replication_delay:d} seconds') def table_is_present_on(context, table_name, pg_name, max_replication_delay): max_replication_delay *= context.timeout_multiplier for _ in range(int(max_replication_delay)): @@ -124,7 +124,7 @@ @step('replication works from {primary:name} to {replica:name} after {time_limit:d} seconds') @then('replication works from {primary:name} to {replica:name} after {time_limit:d} seconds') def replication_works(context, primary, replica, time_limit): - context.execute_steps(u""" + context.execute_steps(""" When I add the table test_{0} to {1} Then table test_{0} is present on {2} after {3} seconds """.format(str(time()).replace('.', '_').replace(',', '_'), primary, replica, time_limit)) diff -Nru patroni-4.0.6/features/steps/cascading_replication.py patroni-4.0.7/features/steps/cascading_replication.py --- patroni-4.0.6/features/steps/cascading_replication.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/features/steps/cascading_replication.py 2025-09-22 18:07:47.000000000 +0200 @@ -9,7 +9,7 @@ return context.pctl.start(name, custom_config={'tags': {tag_name: tag_value}}) -@then('There is a {label} with "{content}" in {name:name} data directory') +@then('there is a {label} with "{content}" in {name:name} data directory') def check_label(context, label, content, name): value = (context.pctl.read_label(name, label) or '').replace('\n', '\\n') assert content in value, "\"{0}\" in {1} doesn't contain {2}".format(value, label, content) diff -Nru patroni-4.0.6/.github/workflows/install_deps.py patroni-4.0.7/.github/workflows/install_deps.py --- patroni-4.0.6/.github/workflows/install_deps.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/.github/workflows/install_deps.py 2025-09-22 18:07:47.000000000 +0200 @@ -10,7 +10,7 @@ def install_requirements(what): subprocess.call([sys.executable, '-m', 'pip', 'install', '--upgrade', 'pip']) - s = subprocess.call([sys.executable, '-m', 'pip', 'install', '--upgrade', 'wheel', 'setuptools']) + s = subprocess.call([sys.executable, '-m', 'pip', 'install', '--upgrade', 'wheel', 'setuptools', 'distlib']) if s != 0: return s @@ -25,11 +25,11 @@ requirements += ['coverage'] # try to split tests between psycopg2 and psycopg3 requirements += ['psycopg[binary]'] if sys.version_info >= (3, 8, 0) and\ - (sys.platform != 'darwin' or what == 'etcd3') else ['psycopg2-binary==2.9.9' + (sys.platform != 'darwin' or what == 'etcd3') else ['psycopg2-binary==2.9.9' if sys.platform == 'darwin' else 'psycopg2-binary'] - from pip._vendor.distlib.markers import evaluator, DEFAULT_CONTEXT - from pip._vendor.distlib.util import parse_requirement + from distlib.markers import evaluator, DEFAULT_CONTEXT + from distlib.util import parse_requirement for r in read('requirements.txt').split('\n'): r = parse_requirement(r) diff -Nru patroni-4.0.6/.github/workflows/release.yaml patroni-4.0.7/.github/workflows/release.yaml --- patroni-4.0.6/.github/workflows/release.yaml 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/.github/workflows/release.yaml 2025-09-22 18:07:47.000000000 +0200 @@ -34,10 +34,10 @@ - name: Publish distribution to Test PyPI if: github.event_name == 'push' - uses: pypa/[email protected] + uses: pypa/[email protected] with: repository_url: https://test.pypi.org/legacy/ - name: Publish distribution to PyPI if: github.event_name == 'release' - uses: pypa/[email protected] + uses: pypa/[email protected] diff -Nru patroni-4.0.6/.github/workflows/tests.yaml patroni-4.0.7/.github/workflows/tests.yaml --- patroni-4.0.6/.github/workflows/tests.yaml 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/.github/workflows/tests.yaml 2025-09-22 18:07:47.000000000 +0200 @@ -198,7 +198,7 @@ - uses: jakebailey/pyright-action@v2 with: - version: 1.1.401 + version: 1.1.405 ydiff: name: Test compatibility with the latest version of ydiff diff -Nru patroni-4.0.6/patroni/api.py patroni-4.0.7/patroni/api.py --- patroni-4.0.6/patroni/api.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/api.py 2025-09-22 18:07:47.000000000 +0200 @@ -517,7 +517,7 @@ ``502`` instead. """ cluster = self.server.patroni.dcs.cluster or self.server.patroni.dcs.get_cluster() - if cluster.config: + if cluster.config and cluster.config.modify_version: self._write_json_response(200, cluster.config.data) else: self.send_error(502) diff -Nru patroni-4.0.6/patroni/dcs/etcd3.py patroni-4.0.7/patroni/dcs/etcd3.py --- patroni-4.0.6/patroni/dcs/etcd3.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/dcs/etcd3.py 2025-09-22 18:07:47.000000000 +0200 @@ -21,7 +21,7 @@ from ..collections import EMPTY_DICT from ..exceptions import DCSError, PatroniException from ..postgresql.mpp import AbstractMPP -from ..utils import deep_compare, enable_keepalive, iter_response_objects, RetryFailedError, USER_AGENT +from ..utils import deep_compare, enable_keepalive, iter_response_objects, parse_bool, RetryFailedError, USER_AGENT from . import catch_return_false_exception, Cluster, ClusterConfig, \ Failover, Leader, Member, Status, SyncState, TimelineHistory from .etcd import AbstractEtcd, AbstractEtcdClientWithFailover, catch_etcd_errors, \ @@ -66,6 +66,10 @@ pass +class Etcd3WatchCanceled(Etcd3Exception): + pass + + class Etcd3ClientError(Etcd3Exception): def __init__(self, code: Optional[int] = None, error: Optional[str] = None, status: Optional[int] = None) -> None: @@ -356,7 +360,6 @@ exc = e self._reauthenticate = True if retry: - logger.error('retry = %s', retry) retry.ensure_deadline(0.5, exc) elif reauthenticated: raise exc @@ -508,6 +511,8 @@ if 'error' in message: raise _raise_for_data(message) result = message.get('result', EMPTY_DICT) + if parse_bool(result.get('canceled')): + raise Etcd3WatchCanceled('Watch canceled') header = result.get('header', EMPTY_DICT) self._check_cluster_raft_term(header.get('cluster_id'), header.get('raft_term')) events: List[Dict[str, Any]] = result.get('events', []) @@ -555,8 +560,10 @@ try: self._do_watch(result['header']['revision']) + except Etcd3WatchCanceled: + logger.info('Watch request canceled') except Exception as e: - # Following exceptions are expected on Windows because the /watch request is done with `read_timeout` + # Following exceptions are expected on Windows because the /watch request is done with `read_timeout` if not (os.name == 'nt' and isinstance(e, (ReadTimeoutError, ProtocolError))): logger.error('watchprefix failed: %r', e) finally: @@ -576,16 +583,21 @@ time.sleep(1) def kill_stream(self) -> None: - sock = None + conn_sock: Any = None with self._response_lock: if isinstance(self._response, urllib3.response.HTTPResponse): try: - sock = self._response.connection.sock if self._response.connection else None + conn_sock = self._response.connection.sock if self._response.connection else None except Exception: - sock = None + conn_sock = None else: self._response = False - if sock: + if conn_sock: + # python-etcd forces usage of pyopenssl if the last one is available. + # In this case HTTPConnection.socket is not inherited from socket.socket, but urllib3 uses custom + # class `WrappedSocket`, which shutdown() method could be incompatible with socket.shutdown(). + # Therefore we use WrappedSocket.socket, which points to original `socket` object. + sock: socket.socket = conn_sock.socket if conn_sock.__class__.__name__ == 'WrappedSocket' else conn_sock try: sock.shutdown(socket.SHUT_RDWR) sock.close() diff -Nru patroni-4.0.6/patroni/dcs/etcd.py patroni-4.0.7/patroni/dcs/etcd.py --- patroni-4.0.6/patroni/dcs/etcd.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/dcs/etcd.py 2025-09-22 18:07:47.000000000 +0200 @@ -317,7 +317,12 @@ # Update machines_cache if previous attempt of update has failed if self._update_machines_cache: - self._load_machines_cache() + try: + self._load_machines_cache() + except etcd.EtcdException as e: + # If etcd cluster isn't accessible _load_machines_cache() -> _refresh_machines_cache() may raise + # etcd.EtcdException. We need to convert it to etcd.EtcdConnectionFailed for failsafe_mode to work. + raise etcd.EtcdConnectionFailed('No more machines in the cluster') from e elif not self._use_proxies and time.time() - self._machines_cache_updated > self._machines_cache_ttl: self._refresh_machines_cache() diff -Nru patroni-4.0.6/patroni/dcs/__init__.py patroni-4.0.7/patroni/dcs/__init__.py --- patroni-4.0.6/patroni/dcs/__init__.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/dcs/__init__.py 2025-09-22 18:07:47.000000000 +0200 @@ -18,7 +18,7 @@ from .. import global_config from ..dynamic_loader import iter_classes, iter_modules -from ..exceptions import PatroniFatalException +from ..exceptions import PatroniAssertionError, PatroniFatalException from ..tags import Tags from ..utils import deep_compare, parse_int, uri @@ -198,10 +198,13 @@ data = {'conn_url': conn_url, 'api_url': api_url} else: try: - data = json.loads(value) - assert isinstance(data, dict) - except (AssertionError, TypeError, ValueError): - data: Dict[str, Any] = {} + json_data = json.loads(value) + if isinstance(json_data, dict): + data = cast(Dict[str, Any], json_data) + else: + raise PatroniAssertionError('not a dict') + except (PatroniAssertionError, TypeError, ValueError): + data = {} return Member(version, name, session, data) @property @@ -480,9 +483,12 @@ data: Dict[str, Any] = value elif value: try: - data = json.loads(value) - assert isinstance(data, dict) - except AssertionError: + json_data = json.loads(value) + if isinstance(json_data, dict): + data = cast(Dict[str, Any], json_data) + else: + raise PatroniAssertionError('not a dict') + except PatroniAssertionError: data = {} except ValueError: t = [a.strip() for a in value.split(':')] @@ -547,10 +553,13 @@ False """ try: - data = json.loads(value) - assert isinstance(data, dict) - except (AssertionError, TypeError, ValueError): - data: Dict[str, Any] = {} + json_data = json.loads(value) + if isinstance(json_data, dict): + data = cast(Dict[str, Any], json_data) + else: + raise PatroniAssertionError('not a dict') + except (PatroniAssertionError, TypeError, ValueError): + data = {} modify_version = 0 return ClusterConfig(version, data, version if modify_version is None else modify_version) @@ -603,11 +612,12 @@ try: if value and isinstance(value, str): value = json.loads(value) - assert isinstance(value, dict) + if not isinstance(value, dict): + raise PatroniAssertionError('not a dict') leader = value.get('leader') quorum = value.get('quorum') return SyncState(version, leader, value.get('sync_standby'), int(quorum) if leader and quorum else 0) - except (AssertionError, TypeError, ValueError): + except (PatroniAssertionError, TypeError, ValueError): return SyncState.empty(version) @staticmethod @@ -737,10 +747,13 @@ [] """ try: - lines = json.loads(value) - assert isinstance(lines, list) - except (AssertionError, TypeError, ValueError): - lines: List[_HistoryTuple] = [] + json_lines = json.loads(value) + if isinstance(json_lines, list): + lines = cast(List[_HistoryTuple], json_lines) + else: + raise PatroniAssertionError('not a list') + except (PatroniAssertionError, TypeError, ValueError): + lines = [] return TimelineHistory(version, value, lines) diff -Nru patroni-4.0.6/patroni/exceptions.py patroni-4.0.7/patroni/exceptions.py --- patroni-4.0.6/patroni/exceptions.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/exceptions.py 2025-09-22 18:07:47.000000000 +0200 @@ -53,3 +53,9 @@ """Any issue identified while loading or validating the Patroni configuration.""" pass + + +class PatroniAssertionError(PatroniException): + """Any issue related to type/value validation.""" + + pass diff -Nru patroni-4.0.6/patroni/__main__.py patroni-4.0.7/patroni/__main__.py --- patroni-4.0.6/patroni/__main__.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/__main__.py 2025-09-22 18:07:47.000000000 +0200 @@ -395,6 +395,12 @@ if pid: os.kill(pid, signo) + import multiprocessing + patroni = multiprocessing.Process(target=patroni_main, args=(args.configfile,)) + patroni.start() + pid = patroni.pid + + # Set up signal handlers after fork to prevent child from inheriting them if os.name != 'nt': signal.signal(signal.SIGCHLD, sigchld_handler) signal.signal(signal.SIGHUP, passtochild) @@ -404,11 +410,6 @@ signal.signal(signal.SIGINT, passtochild) signal.signal(signal.SIGABRT, passtochild) signal.signal(signal.SIGTERM, passtochild) - - import multiprocessing - patroni = multiprocessing.Process(target=patroni_main, args=(args.configfile,)) - patroni.start() - pid = patroni.pid patroni.join() diff -Nru patroni-4.0.6/patroni/postgresql/available_parameters/0_postgres.yml patroni-4.0.7/patroni/postgresql/available_parameters/0_postgres.yml --- patroni-4.0.6/patroni/postgresql/available_parameters/0_postgres.yml 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/postgresql/available_parameters/0_postgres.yml 2025-09-22 18:07:47.000000000 +0200 @@ -121,6 +121,11 @@ version_from: 130000 min_val: -1 max_val: 2147483647 + autovacuum_vacuum_max_threshold: + - type: Integer + version_from: 180000 + min_val: -1 + max_val: 2147483647 autovacuum_vacuum_scale_factor: - type: Real version_from: 90300 @@ -137,6 +142,11 @@ min_val: -1 max_val: 2147483647 unit: kB + autovacuum_worker_slots: + - type: Integer + version_from: 180000 + min_val: 1 + max_val: 262143 backend_flush_after: - type: Integer version_from: 90600 @@ -434,6 +444,9 @@ enable_bitmapscan: - type: Bool version_from: 90300 + enable_distinct_reordering: + - type: Bool + version_from: 180000 enable_gathermerge: - type: Bool version_from: 100000 @@ -485,6 +498,9 @@ enable_presorted_aggregate: - type: Bool version_from: 160000 + enable_self_join_elimination: + - type: Bool + version_from: 180000 enable_seqscan: - type: Bool version_from: 90300 @@ -506,9 +522,13 @@ exit_on_error: - type: Bool version_from: 90300 + extension_control_path: + - type: String + version_from: 180000 extension_destdir: - type: String version_from: 140000 + version_till: 180000 external_pid_file: - type: String version_from: 90300 @@ -517,6 +537,12 @@ version_from: 90300 min_val: -15 max_val: 3 + file_copy_method: + - type: Enum + version_from: 180000 + possible_values: + - copy + - clone force_parallel_mode: - type: EnumBool version_from: 90600 @@ -629,6 +655,12 @@ min_val: 0 max_val: 2147483647 unit: ms + idle_replication_slot_timeout: + - type: Integer + version_from: 180000 + min_val: 0 + max_val: 2147483647 + unit: s idle_session_timeout: - type: Integer version_from: 140000 @@ -647,9 +679,38 @@ io_combine_limit: - type: Integer version_from: 170000 + version_till: 180000 min_val: 1 max_val: 32 unit: 8kB + - type: Integer + version_from: 180000 + min_val: 1 + max_val: 128 + unit: 8kB + io_max_combine_limit: + - type: Integer + version_from: 180000 + min_val: 1 + max_val: 128 + unit: 8kB + io_max_concurrency: + - type: Integer + version_from: 180000 + min_val: -1 + max_val: 1024 + io_method: + - type: Enum + version_from: 180000 + possible_values: + - sync + - worker + - io_uring + io_workers: + - type: Integer + version_from: 180000 + min_val: 1 + max_val: 32 IntervalStyle: - type: Enum version_from: 90300 @@ -748,6 +809,9 @@ log_connections: - type: Bool version_from: 90300 + version_till: 180000 + - type: String + version_from: 180000 log_destination: - type: String version_from: 90300 @@ -793,6 +857,9 @@ log_line_prefix: - type: String version_from: 90300 + log_lock_failures: + - type: Bool + version_from: 180000 log_lock_waits: - type: Bool version_from: 90300 @@ -873,9 +940,15 @@ log_rotation_size: - type: Integer version_from: 90300 + version_till: 180000 min_val: 0 max_val: 2097151 unit: kB + - type: Integer + version_from: 180000 + min_val: 0 + max_val: 2147483647 + unit: kB log_startup_progress_interval: - type: Integer version_from: 150000 @@ -930,9 +1003,20 @@ maintenance_work_mem: - type: Integer version_from: 90300 + version_till: 180000 min_val: 1024 max_val: 2147483647 unit: kB + - type: Integer + version_from: 180000 + min_val: 64 + max_val: 2147483647 + unit: kB + max_active_replication_origins: + - type: Integer + version_from: 180000 + min_val: 0 + max_val: 262143 max_connections: - type: Integer version_from: 90300 @@ -1084,6 +1168,9 @@ version_from: 90600 min_val: 0 max_val: 262143 + md5_password_warnings: + - type: Bool + version_from: 180000 min_dynamic_shared_memory: - type: Integer version_from: 140000 @@ -1139,6 +1226,9 @@ min_val: 16 max_val: 131072 unit: 8kB + oauth_validator_libraries: + - type: String + version_from: 180000 old_snapshot_threshold: - type: Integer version_from: 90600 @@ -1336,6 +1426,10 @@ ssl_ecdh_curve: - type: String version_from: 90400 + version_till: 180000 + ssl_groups: + - type: String + version_from: 180000 ssl_key_file: - type: String version_from: 90300 @@ -1372,6 +1466,9 @@ min_val: 0 max_val: 2147483647 unit: kB + ssl_tls13_ciphers: + - type: String + version_from: 180000 standard_conforming_strings: - type: Bool version_from: 90300 @@ -1547,6 +1644,9 @@ track_commit_timestamp: - type: Bool version_from: 90500 + track_cost_delay_timing: + - type: Bool + version_from: 180000 track_counts: - type: Bool version_from: 90300 @@ -1671,6 +1771,11 @@ version_from: 90300 min_val: 0 max_val: 2000000000 + vacuum_max_eager_freeze_failure_rate: + - type: Real + version_from: 180000 + min_val: 0 + max_val: 1 vacuum_multixact_failsafe_age: - type: Integer version_from: 140000 @@ -1686,6 +1791,9 @@ version_from: 90300 min_val: 0 max_val: 2000000000 + vacuum_truncate: + - type: Bool + version_from: 180000 wal_buffers: - type: Integer version_from: 90300 diff -Nru patroni-4.0.6/patroni/postgresql/config.py patroni-4.0.7/patroni/postgresql/config.py --- patroni-4.0.6/patroni/postgresql/config.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/postgresql/config.py 2025-09-22 18:07:47.000000000 +0200 @@ -1102,8 +1102,11 @@ listen_addresses = self._server_parameters['listen_addresses'].split(',') for la in listen_addresses: - if la.strip().lower() in ('*', '0.0.0.0', '127.0.0.1', 'localhost'): # we are listening on '*' or localhost + if la.strip().lower() in ('*', 'localhost'): # we are listening on '*' or localhost return 'localhost' # connection via localhost is preferred + if la.strip() in ('0.0.0.0', '127.0.0.1'): # Postgres listens only on IPv4 + # localhost, but don't allow Windows to resolve to IPv6 + return '127.0.0.1' if os.name == 'nt' else 'localhost' return listen_addresses[0].strip() # can't use localhost, take first address from listen_addresses def resolve_connection_addresses(self) -> None: diff -Nru patroni-4.0.6/patroni/postgresql/postmaster.py patroni-4.0.7/patroni/postgresql/postmaster.py --- patroni-4.0.6/patroni/postgresql/postmaster.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/postgresql/postmaster.py 2025-09-22 18:07:47.000000000 +0200 @@ -177,7 +177,7 @@ return not self.is_running() def wait_for_user_backends_to_close(self, stop_timeout: Optional[float]) -> None: - # These regexps are cross checked against versions PostgreSQL 9.1 .. 17 + # These regexps are cross checked against versions PostgreSQL 9.1 .. 18 aux_proc_re = re.compile("(?:postgres:)( .*:)? (?:(?:archiver|startup|autovacuum launcher|autovacuum worker|" "checkpointer|logger|stats collector|wal receiver|wal writer|writer)(?: process )?|" "walreceiver|wal sender process|walsender|walwriter|background writer|" @@ -185,7 +185,7 @@ "logical replication tablesync worker for subscription|" "logical replication parallel apply worker for subscription|" "logical replication apply worker for subscription|" - "slotsync worker|walsummarizer|bgworker:) ") + "slotsync worker|walsummarizer|io worker|bgworker:) ") try: children = self.children() diff -Nru patroni-4.0.6/patroni/postgresql/slots.py patroni-4.0.7/patroni/postgresql/slots.py --- patroni-4.0.6/patroni/postgresql/slots.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/postgresql/slots.py 2025-09-22 18:07:47.000000000 +0200 @@ -329,6 +329,26 @@ ' FULL OUTER JOIN dropped ON true'), name) return (rows[0][0], rows[0][1]) if rows else (False, False) + def _drop_replication_slot(self, name: str) -> None: + """Drop replication slot by name. + + .. note:: + If not able to drop the slot, it will log a message and set the flag to reload slots. + + :param name: name of the slot to be dropped. + """ + active, dropped = self.drop_replication_slot(name) + if dropped: + logger.info("Dropped replication slot '%s'", name) + if name in self._replication_slots: + del self._replication_slots[name] + else: + self._schedule_load_slots = True + if active: + logger.warning("Unable to drop replication slot '%s', slot is active", name) + else: + logger.error("Failed to drop replication slot '%s'", name) + def _drop_incorrect_slots(self, cluster: Cluster, slots: Dict[str, Any]) -> None: """Compare required slots and configured as permanent slots with those found, dropping extraneous ones. @@ -344,15 +364,8 @@ # drop old replication slots which are not presented in desired slots. for name in set(self._replication_slots) - set(slots): if not global_config.is_paused and not self.ignore_replication_slot(cluster, name): - active, dropped = self.drop_replication_slot(name) - if dropped: - logger.info("Dropped unknown replication slot '%s'", name) - else: - self._schedule_load_slots = True - if active: - logger.debug("Unable to drop unknown replication slot '%s', slot is still active", name) - else: - logger.error("Failed to drop replication slot '%s'", name) + logger.info("Trying to drop unknown replication slot '%s'", name) + self._drop_replication_slot(name) # drop slots with matching names but attributes that do not match, e.g. `plugin` or `database`. for name, value in slots.items(): @@ -391,15 +404,7 @@ if clean_inactive_physical_slots and value.get('expected_active') is False and value['xmin']: logger.warning('Dropping physical replication slot %s because of its xmin value %s', name, value['xmin']) - active, dropped = self.drop_replication_slot(name) - if dropped: - self._replication_slots.pop(name) - else: - self._schedule_load_slots = True - if active: - logger.warning("Unable to drop replication slot '%s', slot is active", name) - else: - logger.error("Failed to drop replication slot '%s'", name) + self._drop_replication_slot(name) # Now we will create physical replication slots that are missing. if name not in self._replication_slots: @@ -414,8 +419,18 @@ # And advance restart_lsn on physical replication slots that are not expected to be active. elif self._postgresql.can_advance_slots and self._replication_slots[name]['type'] == 'physical' and\ value.get('expected_active') is not True and not value['xmin']: + restart_lsn = value.get('restart_lsn') + if not restart_lsn: + # This slot either belongs to a member or was configured as a permanent slot. However, for some + # reason the slot was created by an external agent instead of by Patroni, and it was created without + # reserving the LSN. We drop the slot here, as we cannot advance it, and let Patroni recreate and + # manage it in the next cycle. + logger.warning('Dropping physical replication slot %s because it has no restart_lsn. ' + 'This slot was probably not created by Patroni, but by an external agent.', name) + self._drop_replication_slot(name) + continue lsn = value.get('lsn') - if lsn and lsn > value['restart_lsn']: # The slot has feedback in DCS and needs to be advanced + if lsn and lsn > restart_lsn: # The slot has feedback in DCS and needs to be advanced try: lsn = format_lsn(lsn) self._query("SELECT pg_catalog.pg_replication_slot_advance(%s, %s)", name, lsn) diff -Nru patroni-4.0.6/patroni/validator.py patroni-4.0.7/patroni/validator.py --- patroni-4.0.6/patroni/validator.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/validator.py 2025-09-22 18:07:47.000000000 +0200 @@ -13,7 +13,7 @@ from .collections import CaseInsensitiveSet, EMPTY_DICT from .dcs import dcs_modules -from .exceptions import ConfigParseError +from .exceptions import ConfigParseError, PatroniAssertionError from .log import type_logformat from .utils import data_directory_is_empty, get_major_version, parse_int, split_host_port @@ -173,9 +173,12 @@ :param value: list of host(s) and port items to be validated. :returns: ``True`` if all items are valid. + + .. note:: + :func:`validate_host_port` will raise an exception if validation failed. """ - assert all([validate_host_port(v) for v in value]), "didn't pass the validation" - return True + + return all(validate_host_port(v) for v in value) def comma_separated_host_port(string: str) -> bool: @@ -870,7 +873,8 @@ :param condition: result of a condition to be asserted. :param message: message to be thrown if the condition is ``False``. """ - assert condition, message + if not condition: + raise PatroniAssertionError(message) class IntValidator(object): diff -Nru patroni-4.0.6/patroni/version.py patroni-4.0.7/patroni/version.py --- patroni-4.0.6/patroni/version.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/patroni/version.py 2025-09-22 18:07:47.000000000 +0200 @@ -2,4 +2,4 @@ :var __version__: the current Patroni version. """ -__version__ = '4.0.6' +__version__ = '4.0.7' diff -Nru patroni-4.0.6/README.rst patroni-4.0.7/README.rst --- patroni-4.0.6/README.rst 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/README.rst 2025-09-22 18:07:47.000000000 +0200 @@ -12,7 +12,7 @@ We call Patroni a "template" because it is far from being a one-size-fits-all or plug-and-play replication system. It will have its own caveats. Use wisely. -Currently supported PostgreSQL versions: 9.3 to 17. +Currently supported PostgreSQL versions: 9.3 to 18. **Note to Citus users**: Starting from 3.0 Patroni nicely integrates with the `Citus <https://github.com/citusdata/citus>`__ database extension to Postgres. Please check the `Citus support page <https://github.com/patroni/patroni/blob/master/docs/citus.rst>`__ in the Patroni documentation for more info about how to use Patroni high availability together with a Citus distributed cluster. diff -Nru patroni-4.0.6/setup.py patroni-4.0.7/setup.py --- patroni-4.0.6/setup.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/setup.py 2025-09-22 18:07:47.000000000 +0200 @@ -35,7 +35,6 @@ 'Environment :: Console', 'Intended Audience :: Developers', 'Intended Audience :: System Administrators', - 'License :: OSI Approved :: MIT License', 'Operating System :: MacOS', 'Operating System :: POSIX :: Linux', 'Operating System :: POSIX :: BSD :: FreeBSD', @@ -205,6 +204,7 @@ author_email=AUTHOR_EMAIL, description=DESCRIPTION, license=LICENSE, + license_files=('LICENSE',), keywords=KEYWORDS, long_description=read('README.rst'), classifiers=CLASSIFIERS, diff -Nru patroni-4.0.6/tests/test_etcd3.py patroni-4.0.7/tests/test_etcd3.py --- patroni-4.0.6/tests/test_etcd3.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/tests/test_etcd3.py 2025-09-22 18:07:47.000000000 +0200 @@ -96,8 +96,12 @@ @patch.object(urllib3.PoolManager, 'urlopen', mock_urlopen) @patch.object(Etcd3Client, 'watchprefix', Mock(return_value=urllib3.response.HTTPResponse())) + @patch.object(urllib3.response.HTTPResponse, 'read_chunked', + Mock(return_value=[b'{"result":{"canceled":true}}'])) def test__build_cache(self): - self.kv_cache._build_cache() + with patch('patroni.dcs.etcd3.logger') as mock_logger: + self.kv_cache._build_cache() + mock_logger.info.assert_called_once_with('Watch request canceled') def test__do_watch(self): self.client.watchprefix = Mock(return_value=False) diff -Nru patroni-4.0.6/tests/test_etcd.py patroni-4.0.7/tests/test_etcd.py --- patroni-4.0.6/tests/test_etcd.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/tests/test_etcd.py 2025-09-22 18:07:47.000000000 +0200 @@ -211,6 +211,9 @@ patch.object(EtcdClient, '_load_machines_cache', Mock(return_value=True)): self.assertRaises(etcd.EtcdException, rtry, self.client.api_execute, '/', 'GET', params={'retry': rtry}) + with patch.object(EtcdClient, '_get_machines_list', Mock(side_effect=etcd.EtcdConnectionFailed)): + self.assertRaises(etcd.EtcdConnectionFailed, self.client.api_execute, '/', 'GET') + with patch.object(EtcdClient, '_do_http_request', Mock(side_effect=etcd.EtcdException)): self.client._read_timeout = 0.01 self.assertRaises(etcd.EtcdException, self.client.api_execute, '/', 'GET') diff -Nru patroni-4.0.6/tests/test_kubernetes.py patroni-4.0.7/tests/test_kubernetes.py --- patroni-4.0.6/tests/test_kubernetes.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/tests/test_kubernetes.py 2025-09-22 18:07:47.000000000 +0200 @@ -23,7 +23,7 @@ def mock_list_namespaced_config_map(*args, **kwargs): k8s_group_label = get_mpp({'citus': {'group': 0, 'database': 'postgres'}}).k8s_group_label metadata = {'resource_version': '1', 'labels': {'f': 'b'}, 'name': 'test-config', - 'annotations': {'initialize': '123', 'config': '{}'}} + 'annotations': {'initialize': '123', 'config': '[]', 'history': '{}'}} items = [k8s_client.V1ConfigMap(metadata=k8s_client.V1ObjectMeta(**metadata))] metadata.update({'name': 'test-leader', 'annotations': {'optime': '1234x', 'leader': 'p-0', 'ttl': '30s', diff -Nru patroni-4.0.6/tests/test_patroni.py patroni-4.0.7/tests/test_patroni.py --- patroni-4.0.6/tests/test_patroni.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/tests/test_patroni.py 2025-09-22 18:07:47.000000000 +0200 @@ -136,9 +136,13 @@ def mock_signal(signo, handler): handler(signo, None) - with patch('signal.signal', mock_signal): - with patch('os.waitpid', Mock(side_effect=[(1, 0), (0, 0)])): + with patch('signal.signal', mock_signal), patch('os.kill') as mock_kill: + with patch('os.waitpid', Mock(side_effect=[(1, 0), (0, 0)])), \ + patch('patroni.__main__.logger') as mock_logger: _main() + mock_kill.assert_called_with(mock_process.return_value.pid, signal.SIGTERM) + if os.name != 'nt': + mock_logger.info.assert_called_with('Reaped pid=%s, exit status=%s', 1, 0) with patch('os.waitpid', Mock(side_effect=OSError)): _main() diff -Nru patroni-4.0.6/tests/test_slots.py patroni-4.0.7/tests/test_slots.py --- patroni-4.0.6/tests/test_slots.py 2025-06-06 19:27:48.000000000 +0200 +++ patroni-4.0.7/tests/test_slots.py 2025-09-22 18:07:47.000000000 +0200 @@ -55,13 +55,13 @@ self.p.set_role('standby_leader') with patch.object(SlotsHandler, 'drop_replication_slot', Mock(return_value=(True, False))), \ patch.object(global_config.__class__, 'is_standby_cluster', PropertyMock(return_value=True)), \ - patch('patroni.postgresql.slots.logger.debug') as mock_debug: + patch('patroni.postgresql.slots.logger.warning') as mock_warning: self.s.sync_replication_slots(cluster, self.tags) - mock_debug.assert_called_once() + mock_warning.assert_called_once_with("Unable to drop replication slot '%s', slot is active", 'foobar') self.p.set_role('replica') with patch.object(Postgresql, 'is_primary', Mock(return_value=False)), \ patch.object(global_config.__class__, 'is_paused', PropertyMock(return_value=True)), \ - patch.object(SlotsHandler, 'drop_replication_slot') as mock_drop: + patch.object(SlotsHandler, '_drop_replication_slot') as mock_drop: config.data['slots'].pop('ls') self.s.sync_replication_slots(cluster, self.tags) mock_drop.assert_not_called() @@ -342,6 +342,7 @@ [self.me, self.other, self.leadermem], None, SyncState.empty(), None, None) global_config.update(cluster) self.s.sync_replication_slots(cluster, self.tags) + with patch.object(SlotsHandler, '_query', Mock(side_effect=[[('blabla', 'physical', None, 12345, None, None, None, None, None)], Exception])) as mock_query, \ patch('patroni.postgresql.slots.logger.error') as mock_error: @@ -353,7 +354,7 @@ with patch.object(SlotsHandler, '_query', Mock(side_effect=[[('test_1', 'physical', 1, 12345, None, None, None, None, None)], Exception])), \ - patch.object(SlotsHandler, 'drop_replication_slot', Mock(return_value=(False, True))): + patch.object(SlotsHandler, '_drop_replication_slot', Mock(return_value=(True))): self.s.sync_replication_slots(cluster, self.tags) with patch.object(SlotsHandler, '_query', Mock(side_effect=[[('test_1', 'physical', 1, 12345, None, None, @@ -366,16 +367,31 @@ with patch.object(SlotsHandler, '_query', Mock(side_effect=[[('test_1', 'physical', 1, 12345, None, None, None, None, None)], Exception])), \ - patch.object(SlotsHandler, 'drop_replication_slot', Mock(return_value=(False, False))): + patch.object(SlotsHandler, '_drop_replication_slot', Mock(return_value=(False))): self.s.sync_replication_slots(cluster, self.tags) with patch.object(SlotsHandler, '_query', Mock(side_effect=[[('test_1', 'physical', 1, 12345, None, None, None, None, None)], Exception])), \ patch.object(Cluster, 'is_unlocked', Mock(return_value=True)), \ - patch.object(SlotsHandler, 'drop_replication_slot') as mock_drop: + patch.object(SlotsHandler, '_drop_replication_slot') as mock_drop: self.s.sync_replication_slots(cluster, self.tags) mock_drop.assert_not_called() + # If the slot has no restart_lsn, we should not try to advance it, and only warn the user that this is not an + # expected situation. + with patch.object(SlotsHandler, '_query', Mock(side_effect=[[('blabla', 'physical', None, None, None, None, + None, None, None)], Exception])) as mock_query, \ + patch('patroni.postgresql.slots.logger.warning') as mock_warning, \ + patch.object(SlotsHandler, '_drop_replication_slot') as mock_drop: + self.s.sync_replication_slots(cluster, self.tags) + for mock_call in mock_query.call_args_list: + self.assertNotIn("pg_catalog.pg_replication_slot_advance", mock_call[0][0]) + self.assertEqual(mock_warning.call_args[0][0], + 'Dropping physical replication slot %s because it has no restart_lsn. ' + 'This slot was probably not created by Patroni, but by an external agent.') + self.assertEqual(mock_warning.call_args[0][1], 'blabla') + mock_drop.assert_called_once_with('blabla') + @patch.object(Postgresql, 'is_primary', Mock(return_value=False)) @patch.object(Postgresql, 'role', PropertyMock(return_value='replica')) @patch.object(TestTags, 'tags', PropertyMock(return_value={'nofailover': True})) @@ -389,3 +405,50 @@ None, None, None)], Exception])) as mock_query: self.s.sync_replication_slots(cluster, self.tags) self.assertTrue(mock_query.call_args[0][0].startswith('SELECT slot_name, slot_type, xmin, ')) + + def test__drop_replication_slot(self): + """Test the :meth:~SlotsHandler._drop_replication_slot` method.""" + # Should log info and remove the slot from the list when the slot is dropped + self.s._replication_slots['testslot'] = {'type': 'physical'} + self.s._schedule_load_slots = False + with patch.object(self.s, 'drop_replication_slot', return_value=(False, True)) as mock_drop, \ + patch('patroni.postgresql.slots.logger.info') as mock_info, \ + patch('patroni.postgresql.slots.logger.warning') as mock_warning, \ + patch('patroni.postgresql.slots.logger.error') as mock_error: + self.s._drop_replication_slot('testslot') + mock_drop.assert_called_once_with('testslot') + mock_info.assert_called_once_with("Dropped replication slot '%s'", 'testslot') + mock_warning.assert_not_called() + mock_error.assert_not_called() + self.assertFalse(self.s._schedule_load_slots) + self.assertNotIn('testslot', self.s._replication_slots) + + # Should log warning and keep slot in the list when the slot is active and not dropped + self.s._replication_slots['testslot'] = {'type': 'physical'} + self.s._schedule_load_slots = False + with patch.object(self.s, 'drop_replication_slot', return_value=(True, False)) as mock_drop, \ + patch('patroni.postgresql.slots.logger.info') as mock_info, \ + patch('patroni.postgresql.slots.logger.warning') as mock_warning, \ + patch('patroni.postgresql.slots.logger.error') as mock_error: + self.s._drop_replication_slot('testslot') + mock_drop.assert_called_once_with('testslot') + mock_info.assert_not_called() + mock_warning.assert_called_once_with("Unable to drop replication slot '%s', slot is active", 'testslot') + mock_error.assert_not_called() + self.assertTrue(self.s._schedule_load_slots) + self.assertIn('testslot', self.s._replication_slots) + + # Should log error and keep the slot in the list when the slot is not active and not dropped + self.s._replication_slots['testslot'] = {'type': 'physical'} + self.s._schedule_load_slots = False + with patch.object(self.s, 'drop_replication_slot', return_value=(False, False)) as mock_drop, \ + patch('patroni.postgresql.slots.logger.info') as mock_info, \ + patch('patroni.postgresql.slots.logger.warning') as mock_warning, \ + patch('patroni.postgresql.slots.logger.error') as mock_error: + self.s._drop_replication_slot('testslot') + mock_drop.assert_called_once_with('testslot') + mock_info.assert_not_called() + mock_warning.assert_not_called() + mock_error.assert_called_once_with("Failed to drop replication slot '%s'", 'testslot') + self.assertTrue(self.s._schedule_load_slots) + self.assertIn('testslot', self.s._replication_slots)
