gitops/patroni
1
0
Fork 0
mirror of https://github.com/outbackdingo/patroni.git synced 2026-09-01 09:09:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

Avoid cloning while bootstrap is running (#2419)

Browse Source 
If cluster has a create replica method that does not require a leader it can get triggered while bootstrap is running. If that method comes up with an accessible cluster faster than the bootstrap completes it will get promoted as the leader, only to lose leader lock when bootstrap completes. To fix this we only consider leaderless create replica methods if sysid is non-empty, i.e. there is no bootstrap running.
This commit is contained in:
Ants Aasma
2022-09-29 13:39:51 +02:00
committed by GitHub
parent 3dcdb16d2a
commit db9b5962ec
2 changed files with 12 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
patroni/ha.py
+3 -1
View File
@@ -273,7 +273,9 @@ class Ha(object):
else:
create_replica_methods = self.get_standby_cluster_config().get('create_replica_methods', []) \
if self.is_standby_cluster() else None
if self.state_handler.can_create_replica_without_replication_connection(create_replica_methods):
can_bootstrap = self.state_handler.can_create_replica_without_replication_connection(create_replica_methods)
concurrent_bootstrap = self.cluster.initialize == ""
if can_bootstrap and not concurrent_bootstrap:
msg = 'bootstrap (without leader)'
return self._async_executor.try_run_async(msg, self.clone) or 'trying to ' + msg
return 'waiting for {0}leader to bootstrap'.format('standby_' if self.is_standby_cluster() else '')
tests/test_ha.py
+9
View File
@@ -45,6 +45,10 @@ def get_cluster_not_initialized_without_leader(cluster_config=None):
return get_cluster(None, None, [], None, SyncState(None, None, None), cluster_config)
def get_cluster_bootstrapping_without_leader(cluster_config=None):
return get_cluster("", None, [], None, SyncState(None, None, None), cluster_config)
def get_cluster_initialized_without_leader(leader=False, failover=None, sync=None, cluster_config=None):
m1 = Member(0, 'leader', 28, {'conn_url': 'postgres://replicator:[email protected]:5435/postgres',
'api_url': 'http://127.0.0.1:8008/patroni', 'xlog_location': 4})
@@ -469,6 +473,11 @@ class TestHa(PostgresInit):
self.p.can_create_replica_without_replication_connection = MagicMock(return_value=True)
self.assertEqual(self.ha.bootstrap(), 'trying to bootstrap (without leader)')
def test_bootstrap_not_running_concurrently(self):
self.ha.cluster = get_cluster_bootstrapping_without_leader()
self.p.can_create_replica_without_replication_connection = MagicMock(return_value=True)
self.assertEqual(self.ha.bootstrap(), 'waiting for leader to bootstrap')
def test_bootstrap_initialize_lock_failed(self):
self.ha.cluster = get_cluster_not_initialized_without_leader()
self.assertEqual(self.ha.bootstrap(), 'failed to acquire initialize lock')