Skip to content

fix switch over candidate retrieving #2760

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Nov 1, 2024
Merged

fix switch over candidate retrieving #2760

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
ce91e5f
fix switch over candidate retrieving
Sep 12, 2024
97edacf
Update pkg/cluster/pod.go
FxKu Nov 1, 2024
c3a9cf2
Update pkg/cluster/pod.go
FxKu Nov 1, 2024
ef9c4cb
Merge branch 'master' into fix-asynchronous-get-switchover-candidates…
FxKu Nov 1, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
33 changes: 15 additions & 18 deletions pkg/cluster/pod.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -480,6 +480,9 @@ func (c *Cluster) getSwitchoverCandidate(master *v1.Pod) (spec.NamespacedName, e
if PostgresRole(member.Role) == SyncStandby {
syncCandidates = append(syncCandidates, member)
}
if PostgresRole(member.Role) != Leader && PostgresRole(member.Role) != StandbyLeader && slices.Contains([]string{"running", "streaming", "in archive recovery"}, member.State) {
candidates = append(candidates, member)
}
}

// if synchronous mode is enabled and no SyncStandy was found
Expand All @@ -489,6 +492,12 @@ func (c *Cluster) getSwitchoverCandidate(master *v1.Pod) (spec.NamespacedName, e
return false, nil
}

// retry also in asynchronous mode when no replica candidate was found
if !c.Spec.Patroni.SynchronousMode && len(candidates) == 0 {
c.logger.Warnf("no replica candidate found - retrying fetching cluster members")
return false, nil
}

return true, nil
},
)
Expand All @@ -502,24 +511,12 @@ func (c *Cluster) getSwitchoverCandidate(master *v1.Pod) (spec.NamespacedName, e
return syncCandidates[i].Lag < syncCandidates[j].Lag
})
return spec.NamespacedName{Namespace: master.Namespace, Name: syncCandidates[0].Name}, nil
} else {
// in asynchronous mode find running replicas
for _, member := range members {
if PostgresRole(member.Role) == Leader || PostgresRole(member.Role) == StandbyLeader {
continue
}

if slices.Contains([]string{"running", "streaming", "in archive recovery"}, member.State) {
candidates = append(candidates, member)
}
}

if len(candidates) > 0 {
sort.Slice(candidates, func(i, j int) bool {
return candidates[i].Lag < candidates[j].Lag
})
return spec.NamespacedName{Namespace: master.Namespace, Name: candidates[0].Name}, nil
}
}
if len(candidates) > 0 {
sort.Slice(candidates, func(i, j int) bool {
return candidates[i].Lag < candidates[j].Lag
})
return spec.NamespacedName{Namespace: master.Namespace, Name: candidates[0].Name}, nil
}

return spec.NamespacedName{}, fmt.Errorf("no switchover candidate found")
Expand Down
4 changes: 2 additions & 2 deletions pkg/cluster/pod_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -62,7 +62,7 @@ func TestGetSwitchoverCandidate(t *testing.T) {
expectedError: nil,
},
{
subtest: "choose first replica when lag is equal evrywhere",
subtest: "choose first replica when lag is equal everywhere",
clusterJson: `{"members": [{"name": "acid-test-cluster-0", "role": "leader", "state": "running", "api_url": "http://192.168.100.1:8008/patroni", "host": "192.168.100.1", "port": 5432, "timeline": 1}, {"name": "acid-test-cluster-1", "role": "replica", "state": "streaming", "api_url": "http://192.168.100.2:8008/patroni", "host": "192.168.100.2", "port": 5432, "timeline": 1, "lag": 5}, {"name": "acid-test-cluster-2", "role": "replica", "state": "running", "api_url": "http://192.168.100.3:8008/patroni", "host": "192.168.100.3", "port": 5432, "timeline": 1, "lag": 5}]}`,
syncModeEnabled: false,
expectedCandidate: spec.NamespacedName{Namespace: namespace, Name: "acid-test-cluster-1"},
Expand All @@ -73,7 +73,7 @@ func TestGetSwitchoverCandidate(t *testing.T) {
clusterJson: `{"members": [{"name": "acid-test-cluster-0", "role": "leader", "state": "running", "api_url": "http://192.168.100.1:8008/patroni", "host": "192.168.100.1", "port": 5432, "timeline": 2}, {"name": "acid-test-cluster-1", "role": "replica", "state": "starting", "api_url": "http://192.168.100.2:8008/patroni", "host": "192.168.100.2", "port": 5432, "timeline": 2}]}`,
syncModeEnabled: false,
expectedCandidate: spec.NamespacedName{},
expectedError: fmt.Errorf("no switchover candidate found"),
expectedError: fmt.Errorf("failed to get Patroni cluster members: unexpected end of JSON input"),
},
{
subtest: "replicas with different status",
Expand Down
Loading