Compare commits

...
6 Commits
Author SHA1 Message Date
Alexander Kukushkin 710afd5952 Release v3.1.2 (#2885)
- bump version
- update release notes
2023-09-26 12:31:18 +02:00
Alexander Kukushkin 096ee8f36f Read GUC's values when joining running Postgres (#2876)
If restarted in pause Patroni was discarding `synchronous_standby_names` from `postgresql.conf` because in the internal cache this values was set to `None`. As a result synchronous replication transitioned to a broken state, with no synchronous replicas according to the `synchronous_standby_names` and Patroni not selecting/setting the new synchronous replicas (another bug).

To solve the problem of broken initial state and to avoid similar issues with other GUC's we will read GUC's value if Patroni is joining running Postgres.
2023-09-26 11:34:19 +02:00
Alexander Kukushkin 91e2be092c Detect and solve inconsistency between /sync and actual sync nodes (#2877)
Patroni is changing `synchronous_standby_names` and the `/sync` key in a very specific order, first we add nodes to `synchronous_standby_names` and only after, when they are recognized as synchronous they are added to the `/sync` key. When removing nodes the order is different: they are first removed from the `/sync` key and only after that from the `synchronous_standby_names`.

As a result Patroni expects that either actual synchronous nodes will match with the nodes listed in the `/sync` key or that new candidates to synchronous nodes will not match with nodes listed in the `/sync` key. In case if `synchronous_standby_names` was removed from the `postgresql.conf`, manually, or due the the bug (#2876), the state becomes inconsistent because of the wrong order of updates.

To solve inconsistent state we introduce additional checks and will update the `/sync` key with actual names of synchronous nodes (usually empty set).
2023-09-26 11:17:12 +02:00
Alexander Kukushkin 4148e0b5b2 Take into account current role when deciding on removal of member ZNode (#2884)
Patroni doesn't watch on all changes of member keys in order to not create too much load on ZooKeeper, but only subscribes to changes (ZNodes added or deleted) in the `/member` directory. Therefore when some important fields in the value are updated we remove and recreate ZNode in order to notify the leader or other members.

The leader should remove the member key only when the `checkpoint_after_promote` value is changed and replicas when the `state` is changed to/from `running`.

We don't care about the `version` field, because Patroni version can't be changed without restart, what will case ZooKeeper `session_id` to change it anyway.

This fix hopefully will reduce failures of behave tests on GH Actions.
2023-09-26 11:16:52 +02:00
Alexander Kukushkin 5ceba81269 Bugfix for GUC's values with units (#2883)
Despite being validated by `IntValidator` some GUC's couldn't be casted directly to `int` because they include suffix. Example: `128MB`.

Close https://github.com/zalando/patroni/issues/2879
2023-09-26 11:16:47 +02:00
Alexander Kukushkin dbbe065a27 Silence annoying warnings when checking for node uniqueness (#2878)
WARNING messages are produced by `urllib3` if Patroni is quickly restarted.
Instead we will check that the node is listen on a given port. This fact is actually enough to detect names clashes, while HTTP request could raise an exception is a few other cases, what might case false negatives.

Close https://github.com/zalando/patroni/issues/2881
2023-09-26 11:16:38 +02:00
15 changed files with 145 additions and 38 deletions
+22
View File
@@ -3,6 +3,28 @@
Release notes
=============
Version 3.1.2
-------------
**Bugfixes**
- Fixed bug with ``wal_keep_size`` checks (Alexander Kukushkin)
The ``wal_keep_size`` is a GUC that normally has a unit and Patroni was failing to cast its value to ``int``. As a result the value of ``bootstrap.dcs`` was not written to the ``/config`` key afterwards.
- Detect and resolve inconsistencies between ``/sync`` key and ``synchronous_standby_names`` (Alexander Kukushkin)
Normally, Patroni updates ``/sync`` and ``synchronous_standby_names`` in a very specific order, but in case of a bug or when someone manually reset ``synchronous_standby_names``, Patroni was getting into an inconsistent state. As a result it was possible that the failover happens to an asynchronous node.
- Read GUC's values when joining running Postgres (Alexander Kukushkin)
When restarted in ``pause``, Patroni was discarding the ``synchronous_standby_names`` GUC from the ``postgresql.conf``. To solve it and avoid similar issues, Patroni will read GUC's value if it is joining an already running Postgres.
- Silenced annoying warnings when checking for node uniqueness (Alexander Kukushkin)
``WARNING`` messages are produced by ``urllib3`` if Patroni is quickly restarted.
Version 3.1.1
-------------
+2 -1
View File
@@ -68,6 +68,7 @@ Scenario: check API requests for the primary-replica pair in the pause mode
When I kill postmaster on postgres1
And I issue a GET request to http://127.0.0.1:8009/replica
Then I receive a response code 503
And "members/postgres1" key in DCS has state=stopped after 10 seconds
When I run patronictl.py restart batman postgres1 --force
Then I receive a response returncode 0
Then replication works from postgres0 to postgres1 after 20 seconds
@@ -76,7 +77,7 @@ Scenario: check API requests for the primary-replica pair in the pause mode
Then I receive a response code 200
And I receive a response state running
And I receive a response role replica
When I run patronictl.py reinit batman postgres1 --force
When I run patronictl.py reinit batman postgres1 --force --wait
Then I receive a response returncode 0
And I receive a response output "Success: reinitialize for member postgres1"
And postgres1 role is the secondary after 30 seconds
+8 -3
View File
@@ -65,6 +65,8 @@ class Patroni(AbstractPatroniDaemon):
def ensure_unique_name(self) -> None:
"""A helper method to prevent splitbrain from operator naming error."""
from urllib.parse import urlparse
from urllib3.connection import HTTPConnection
from patroni.dcs import Member
cluster = self.dcs.get_cluster()
@@ -74,9 +76,12 @@ class Patroni(AbstractPatroniDaemon):
if not isinstance(member, Member):
return
try:
_ = self.request(member, endpoint="/liveness", timeout=3)
logger.fatal("Can't start; there is already a node named '%s' running", self.config['name'])
sys.exit(1)
parts = urlparse(member.api_url)
if isinstance(parts.hostname, str):
connection = HTTPConnection(parts.hostname, port=parts.port or 80, timeout=3)
connection.connect()
logger.fatal("Can't start; there is already a node named '%s' running", self.config['name'])
sys.exit(1)
except Exception:
return
+2 -1
View File
@@ -342,7 +342,8 @@ class Config(object):
elif not is_local:
validator = ConfigHandler.CMDLINE_OPTIONS[name][1]
if validator(value):
pg_params[name] = int(value) if isinstance(validator, IntValidator) else value
int_val = parse_int(value) if isinstance(validator, IntValidator) else None
pg_params[name] = int_val if isinstance(int_val, int) else value
else:
logger.warning("postgresql parameter %s=%s failed validation, defaulting to %s",
name, value, ConfigHandler.CMDLINE_OPTIONS[name][0])
+22 -15
View File
@@ -393,21 +393,28 @@ class ZooKeeper(AbstractDCS):
cluster = self.cluster
member = cluster and cluster.get_member(self._name, fallback_to_leader=False)
member_data = self.__last_member_data or member and member.data
# We want to notify leader if some important fields in the member key changed by removing ZNode
if member and (self._client.client_id is not None and member.session != self._client.client_id[0]
or not (member_data and deep_compare(member_data.get('tags', {}), data.get('tags', {}))
and (member_data.get('state') == data.get('state')
or 'running' not in (member_data.get('state'), data.get('state')))
and member_data.get('version') == data.get('version')
and member_data.get('checkpoint_after_promote')
== data.get('checkpoint_after_promote'))):
try:
self._client.delete_async(self.member_path).get(timeout=1)
except NoNodeError:
pass
except Exception:
return False
member = None
if member and member_data:
is_leader = data.get('role') in ('master', 'primary', 'standby_leader')
checkpoint_after_promote_changed = member_data.get('checkpoint_after_promote') \
!= data.get('checkpoint_after_promote')
state_running_changed = member_data.get('state') != data.get('state') \
and 'running' in (member_data.get('state'), data.get('state'))
tags_changed = not deep_compare(member_data.get('tags', {}), data.get('tags', {}))
# We want delete the member ZNode if:
# - our session doesn't match with session id on our member key; or
# - we want to notify leader if some important fields in the member key changed; or
# - if we are the leader and want to notify replicas about checkpoint_after_promote;
if self._client.client_id is not None and member.session != self._client.client_id[0] \
or is_leader and checkpoint_after_promote_changed \
or not is_leader and (state_running_changed or tags_changed):
try:
self._client.delete_async(self.member_path).get(timeout=1)
except NoNodeError:
pass
except Exception:
return False
member = None
encoded_data = json.dumps(data, separators=(',', ':')).encode('utf-8')
if member and member_data:
+8
View File
@@ -654,6 +654,14 @@ class Ha(object):
current = CaseInsensitiveSet(sync.members)
picked, allow_promote = self.state_handler.sync_handler.current_state(self.cluster)
if picked == current and current != allow_promote:
logger.warning('Inconsistent state between synchronous_standby_names = %s and /sync = %s key '
'detected, updating synchronous replication key...', list(allow_promote), list(current))
sync = self.dcs.write_sync_state(self.state_handler.name, allow_promote, version=sync.version)
if not sync:
return logger.warning("Updating sync state failed")
current = CaseInsensitiveSet(sync.members)
if picked != current:
# update synchronous standby list in dcs temporarily to point to common nodes in current and picked
sync_common = current & allow_promote
+18 -7
View File
@@ -118,17 +118,28 @@ class Postgresql(object):
# Last known running process
self._postmaster_proc = None
if self.is_running(): # we are "joining" already running postgres
self.set_state('running')
if self.is_running():
# If we found postmaster process we need to figure out whether postgres is accepting connections
self.set_state('starting')
self.check_startup_state_changed()
if self.state == 'running': # we are "joining" already running postgres
# we know that PostgreSQL is accepting connections and can read some GUC's from pg_settings
self.config.load_current_server_parameters()
self.set_role('master' if self.is_leader() else 'replica')
# postpone writing postgresql.conf for 12+ because recovery parameters are not yet known
if self.major_version < 120000 or self.is_leader():
self.config.write_postgresql_conf()
hba_saved = self.config.replace_pg_hba()
ident_saved = self.config.replace_pg_ident()
if hba_saved or ident_saved:
if self.major_version < 120000 or self.role in ('master', 'primary'):
# If PostgreSQL is running as a primary or we run PostgreSQL that is older than 12 we can
# call reload_config() once again (the first call happened in the ConfigHandler constructor),
# so that it can figure out if config files should be updated and pg_ctl reload executed.
self.config.reload_config(config, sighup=bool(hba_saved or ident_saved))
elif hba_saved or ident_saved:
self.reload()
elif self.role in ('master', 'primary'):
elif not self.is_running() and self.role in ('master', 'primary'):
self.set_role('demoted')
@property
+14 -6
View File
@@ -326,14 +326,22 @@ class ConfigHandler(object):
.format(self._pgpass))
self._passfile = None
self._passfile_mtime = None
self._synchronous_standby_names = None
self._postmaster_ctime = None
self._current_recovery_params: Optional[CaseInsensitiveDict] = None
self._config = {}
self._recovery_params = CaseInsensitiveDict()
self._server_parameters: CaseInsensitiveDict
self._server_parameters: CaseInsensitiveDict = CaseInsensitiveDict()
self.reload_config(config)
def load_current_server_parameters(self) -> None:
"""Read GUC's values from ``pg_settings`` when Patroni is joining the the postgres that is already running."""
exclude = [name.lower() for name, value in self.CMDLINE_OPTIONS.items() if value[1] == _false_validator] \
+ [name.lower() for name in self._RECOVERY_PARAMETERS]
self._server_parameters = CaseInsensitiveDict({r[0]: r[1] for r in self._postgresql.query(
"SELECT name, pg_catalog.current_setting(name) FROM pg_catalog.pg_settings"
" WHERE (source IN ('command line', 'environment variable') OR sourcefile = %s)"
" AND pg_catalog.lower(name) != ALL(%s)", self._postgresql_conf, exclude)})
def setup_server_parameters(self) -> None:
self._server_parameters = self.get_server_parameters(self._config)
self._adjust_recovery_parameters()
@@ -922,14 +930,15 @@ class ConfigHandler(object):
listen_addresses, port = split_host_port(config['listen'], 5432)
parameters.update(cluster_name=self._postgresql.scope, listen_addresses=listen_addresses, port=str(port))
if not self._postgresql.global_config or self._postgresql.global_config.is_synchronous_mode:
if self._synchronous_standby_names is None:
synchronous_standby_names = self._server_parameters.get('synchronous_standby_names')
if synchronous_standby_names is None:
if self._postgresql.global_config and self._postgresql.global_config.is_synchronous_mode_strict\
and self._postgresql.role in ('master', 'primary', 'promoted'):
parameters['synchronous_standby_names'] = '*'
else:
parameters.pop('synchronous_standby_names', None)
else:
parameters['synchronous_standby_names'] = self._synchronous_standby_names
parameters['synchronous_standby_names'] = synchronous_standby_names
# Handle hot_standby <-> replica rename
if parameters.get('wal_level') == ('hot_standby' if self._postgresql.major_version >= 90600 else 'replica'):
@@ -1129,12 +1138,11 @@ class ConfigHandler(object):
def set_synchronous_standby_names(self, value: Optional[str]) -> Optional[bool]:
"""Updates synchronous_standby_names and reloads if necessary.
:returns: True if value was updated."""
if value != self._synchronous_standby_names:
if value != self._server_parameters.get('synchronous_standby_names'):
if value is None:
self._server_parameters.pop('synchronous_standby_names', None)
else:
self._server_parameters['synchronous_standby_names'] = value
self._synchronous_standby_names = value
if self._postgresql.state == 'running':
self.write_postgresql_conf()
self._postgresql.reload()
+1 -1
View File
@@ -2,4 +2,4 @@
:var __version__: the current Patroni version.
"""
__version__ = '3.1.1'
__version__ = '3.1.2'
+21
View File
@@ -118,11 +118,31 @@ class MockCursor(object):
'"state":"streaming","sync_state":"async","sync_priority":0}]'
now = datetime.datetime.now(tzutc)
self.results = [(now, 0, '', 0, '', False, now, 'streaming', None, replication_info)]
elif sql.startswith('SELECT name, pg_catalog.current_setting(name) FROM pg_catalog.pg_settings'):
self.results = [('data_directory', 'data'),
('hba_file', os.path.join('data', 'pg_hba.conf')),
('ident_file', os.path.join('data', 'pg_ident.conf')),
('max_connections', 42),
('max_locks_per_transaction', 73),
('max_prepared_transactions', 0),
('max_replication_slots', 21),
('max_wal_senders', 37),
('track_commit_timestamp', 'off'),
('wal_level', 'replica'),
('listen_addresses', '6.6.6.6'),
('port', 1984),
('archive_command', 'my archive command'),
('cluster_name', 'my_cluster')]
elif sql.startswith('SELECT name, setting'):
self.results = [('wal_segment_size', '2048', '8kB', 'integer', 'internal'),
('wal_block_size', '8192', None, 'integer', 'internal'),
('shared_buffers', '16384', '8kB', 'integer', 'postmaster'),
('wal_buffers', '-1', '8kB', 'integer', 'postmaster'),
('max_connections', '100', None, 'integer', 'postmaster'),
('max_prepared_transactions', '0', None, 'integer', 'postmaster'),
('max_worker_processes', '8', None, 'integer', 'postmaster'),
('max_locks_per_transaction', '64', None, 'integer', 'postmaster'),
('max_wal_senders', '5', None, 'integer', 'postmaster'),
('search_path', 'public', None, 'string', 'user'),
('port', '5433', None, 'integer', 'postmaster'),
('listen_addresses', '*', None, 'string', 'postmaster'),
@@ -222,6 +242,7 @@ class PostgresInit(unittest.TestCase):
class BaseTestPostgresql(PostgresInit):
@patch('time.sleep', Mock())
def setUp(self):
super(BaseTestPostgresql, self).setUp()
+1
View File
@@ -155,6 +155,7 @@ class TestConfig(unittest.TestCase):
expected_params = {
'f.oo': 'bar', # not in ConfigHandler.CMDLINE_OPTIONS
'max_connections': 100, # IntValidator
'wal_keep_size': '128MB', # IntValidator
'wal_level': 'hot_standby', # EnumValidator
}
input_params = deepcopy(expected_params)
+18
View File
@@ -1313,6 +1313,24 @@ class TestHa(PostgresInit):
self.ha.run_cycle()
self.assertEqual(mock_logger.call_args[0][0], 'Updating sync state failed')
@patch.object(Cluster, 'is_unlocked', Mock(return_value=False))
def test_inconsistent_synchronous_state(self):
self.ha.is_synchronous_mode = true
self.ha.has_lock = true
self.p.name = 'leader'
self.ha.cluster = get_cluster_initialized_without_leader(sync=('leader', 'a'))
self.p.sync_handler.current_state = Mock(return_value=(CaseInsensitiveSet('a'), CaseInsensitiveSet()))
self.ha.dcs.write_sync_state = Mock(return_value=SyncState.empty())
mock_set_sync = self.p.sync_handler.set_synchronous_standby_names = Mock()
with patch('patroni.ha.logger.warning') as mock_logger:
self.ha.run_cycle()
mock_set_sync.assert_called_once()
self.assertTrue(mock_logger.call_args_list[0][0][0].startswith('Inconsistent state between '))
self.ha.dcs.write_sync_state = Mock(return_value=None)
with patch('patroni.ha.logger.warning') as mock_logger:
self.ha.run_cycle()
self.assertEqual(mock_logger.call_args[0][0], 'Updating sync state failed')
def test_effective_tags(self):
self.ha._disable_sync = True
self.assertEqual(self.ha.get_effective_tags(), {'foo': 'bar', 'nosync': True})
+5 -4
View File
@@ -40,7 +40,7 @@ class MockFrozenImporter(object):
@patch('time.sleep', Mock())
@patch('subprocess.call', Mock(return_value=0))
@patch('patroni.psycopg.connect', psycopg_connect)
@patch('urllib3.PoolManager.request', Mock(side_effect=Exception))
@patch('urllib3.connection.HTTPConnection.connect', Mock(side_effect=Exception))
@patch.object(ConfigHandler, 'append_pg_hba', Mock())
@patch.object(ConfigHandler, 'write_postgresql_conf', Mock())
@patch.object(ConfigHandler, 'write_recovery_conf', Mock())
@@ -64,7 +64,7 @@ class TestPatroni(unittest.TestCase):
self.assertRaises(SystemExit, _main)
@patch('pkgutil.iter_importers', Mock(return_value=[MockFrozenImporter()]))
@patch('urllib3.PoolManager.request', Mock(side_effect=Exception))
@patch('urllib3.connection.HTTPConnection.connect', Mock(side_effect=Exception))
@patch('sys.frozen', Mock(return_value=True), create=True)
@patch.object(HTTPServer, '__init__', Mock())
@patch.object(etcd.Client, 'read', etcd_read)
@@ -108,6 +108,7 @@ class TestPatroni(unittest.TestCase):
@patch('os.getpid')
@patch('multiprocessing.Process')
@patch('patroni.__main__.patroni_main', Mock())
@patch('sys.argv', ['patroni.py', 'postgres0.yml'])
def test_patroni_main(self, mock_process, mock_getpid):
mock_getpid.return_value = 2
_main()
@@ -233,8 +234,8 @@ class TestPatroni(unittest.TestCase):
)
with patch('patroni.dcs.AbstractDCS.get_cluster', Mock(return_value=bad_cluster)):
# If the api of the running node cannot be reached, this implies unique name
with patch.object(self.p, 'request', Mock(side_effect=ConnectionError)):
with patch('urllib3.connection.HTTPConnection.connect', Mock(side_effect=ConnectionError)):
self.assertIsNone(self.p.ensure_unique_name())
# Only if the api of the running node is reachable do we throw an error
with patch.object(self.p, 'request', Mock()):
with patch('urllib3.connection.HTTPConnection.connect', Mock()):
self.assertRaises(SystemExit, self.p.ensure_unique_name)
+2
View File
@@ -715,6 +715,7 @@ class TestPostgresql(BaseTestPostgresql):
self.assertEqual(self.p.get_primary_timeline(), 1)
@patch.object(Postgresql, 'get_postgres_role_from_data_directory', Mock(return_value='replica'))
@patch.object(Postgresql, 'is_running', Mock(return_value=False))
@patch.object(Bootstrap, 'running_custom_bootstrap', PropertyMock(return_value=True))
@patch.object(Postgresql, 'controldata', Mock(return_value={'max_connections setting': '200',
'max_worker_processes setting': '20',
@@ -958,6 +959,7 @@ class TestPostgresql2(BaseTestPostgresql):
@patch('patroni.postgresql.CallbackExecutor', Mock())
@patch.object(Postgresql, 'get_major_version', Mock(return_value=140000))
@patch.object(Postgresql, 'is_running', Mock(return_value=True))
@patch.object(Postgresql, 'is_leader', Mock(return_value=False))
def setUp(self):
super(TestPostgresql2, self).setUp()
+1
View File
@@ -276,6 +276,7 @@ class TestZooKeeper(unittest.TestCase):
self.assertTrue(self.zk.delete_cluster())
def test_watch(self):
self.zk.event.wait = Mock()
self.zk.watch(None, 0)
self.zk.event.is_set = Mock(return_value=True)
self.zk._fetch_status = False