Skip to content

fix(slots): cleanup replication slots on replicas when features disabled #9381

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
leonardoce merged 4 commits into from
Dec 15, 2025
Merged

fix(slots): cleanup replication slots on replicas when features disabled #9381

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
4ebd36e
fix(slots): cleanup replication slots on replicas when features disabled
mnencia Dec 7, 2025
82fa94e
fix: webhook validation
armru Dec 10, 2025
3198161
tests: improve coverage
mnencia Dec 11, 2025
5d86138
chore: spelling and grammar
mnencia Dec 11, 2025
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
11 changes: 8 additions & 3 deletions internal/management/controller/slots/runner/runner.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,13 +72,18 @@ func (sr *Replicator) Start(ctx context.Context) error {
case <-ticker.C:
}

// If replication is disabled stop the timer,
// the process will resume through the wakeUp channel if necessary
// If replication is disabled, stop the timer and perform cleanup
if config == nil || !config.GetEnabled() {
ticker.Stop()
// we set updateInterval to 0 to make sure the Ticker will be reset
// We set updateInterval to 0 to make sure the Ticker will be reset
// if the feature is enabled again
updateInterval = 0
// Perform cleanup before continuing
if config != nil {
if err := sr.reconcile(ctx, config); err != nil {
contextLog.Warning("cleanup during reconcile when features are disabled", "err", err)
}
}
continue
}

Expand Down
65 changes: 43 additions & 22 deletions internal/webhook/v1/cluster_webhook.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2273,33 +2273,54 @@ func (v *ClusterCustomValidator) validateReplicationSlotsChange(r, old *apiv1.Cl
newReplicationSlots := r.Spec.ReplicationSlots
oldReplicationSlots := old.Spec.ReplicationSlots

if oldReplicationSlots == nil || oldReplicationSlots.HighAvailability == nil ||
!oldReplicationSlots.HighAvailability.GetEnabled() {
var errs field.ErrorList

if oldReplicationSlots == nil {
return nil
}

var errs field.ErrorList
// Validate HighAvailability changes
if oldReplicationSlots.HighAvailability.GetEnabled() {
// When disabling, we should check that the prefix is not removed and doesn't change
// to properly execute the cleanup logic
if newReplicationSlots == nil || newReplicationSlots.HighAvailability == nil {
path := field.NewPath("spec", "replicationSlots")
if newReplicationSlots != nil {
path = path.Child("highAvailability")
}
errs = append(errs,
field.Invalid(
path,
nil,
fmt.Sprintf("Cannot remove %v section while highAvailability is enabled", path)),
)
} else if oldReplicationSlots.HighAvailability.SlotPrefix != newReplicationSlots.HighAvailability.SlotPrefix {
errs = append(errs,
field.Invalid(
field.NewPath("spec", "replicationSlots", "highAvailability", "slotPrefix"),
newReplicationSlots.HighAvailability.SlotPrefix,
"Cannot change replication slot prefix while highAvailability is enabled"),
)
}
}

// when disabling we should check that the prefix it's not removed, and it doesn't change to
// properly execute the cleanup logic
if newReplicationSlots == nil || newReplicationSlots.HighAvailability == nil {
path := field.NewPath("spec", "replicationSlots")
if newReplicationSlots != nil {
path = path.Child("highAvailability")
// Validate SynchronizeReplicas changes
// When synchronizeReplicas is enabled, we need to ensure users disable it before removing the configuration
// to allow the cleanup logic to properly remove user-defined replication slots from replicas
if oldReplicationSlots.SynchronizeReplicas.GetEnabled() {
if newReplicationSlots == nil || newReplicationSlots.SynchronizeReplicas == nil {
path := field.NewPath("spec", "replicationSlots")
if newReplicationSlots != nil {
path = path.Child("synchronizeReplicas")
}
errs = append(errs,
field.Invalid(
path,
nil,
fmt.Sprintf("Cannot remove %v section while synchronizeReplicas is enabled. "+
"Disable synchronizeReplicas first to allow cleanup of user-defined replication slots on replicas", path)),
)
}
errs = append(errs,
field.Invalid(
path,
nil,
fmt.Sprintf("Cannot remove %v section while highAvailability is enabled", path)),
)
} else if oldReplicationSlots.HighAvailability.SlotPrefix != newReplicationSlots.HighAvailability.SlotPrefix {
errs = append(errs,
field.Invalid(
field.NewPath("spec", "replicationSlots", "highAvailability", "slotPrefix"),
newReplicationSlots.HighAvailability.SlotPrefix,
"Cannot change replication slot prefix while highAvailability is enabled"),
)
}

return errs
Expand Down
120 changes: 119 additions & 1 deletion internal/webhook/v1/cluster_webhook_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3546,6 +3546,9 @@ var _ = Describe("validation of replication slots configuration", func() {
HighAvailability: &apiv1.ReplicationSlotsHAConfiguration{
Enabled: ptr.To(false),
},
SynchronizeReplicas: &apiv1.SynchronizeReplicasConfiguration{
Enabled: ptr.To(false),
},
},
},
}
Expand All @@ -3557,6 +3560,9 @@ var _ = Describe("validation of replication slots configuration", func() {
Enabled: ptr.To(true),
SlotPrefix: "_test_",
},
SynchronizeReplicas: &apiv1.SynchronizeReplicasConfiguration{
Enabled: ptr.To(true),
},
}

Expect(v.validateReplicationSlotsChange(newCluster, oldCluster)).To(BeEmpty())
Expand All @@ -3581,7 +3587,7 @@ var _ = Describe("validation of replication slots configuration", func() {
Expect(v.validateReplicationSlotsChange(newCluster, oldCluster)).To(HaveLen(1))
})

It("prevents removing the replication slot section when replication slots are enabled", func() {
It("prevents removing the replication slot section when highAvailability is enabled", func() {
oldCluster := &apiv1.Cluster{
Spec: apiv1.ClusterSpec{
ImageName: versions.DefaultImageName,
Expand All @@ -3590,6 +3596,30 @@ var _ = Describe("validation of replication slots configuration", func() {
Enabled: ptr.To(true),
SlotPrefix: "_test_",
},
SynchronizeReplicas: &apiv1.SynchronizeReplicasConfiguration{
Enabled: ptr.To(false),
},
},
},
}
oldCluster.Default()

newCluster := oldCluster.DeepCopy()
newCluster.Spec.ReplicationSlots = nil
Expect(v.validateReplicationSlotsChange(newCluster, oldCluster)).To(HaveLen(1))
})

It("prevents removing the replication slot section when synchronizeReplicas is enabled", func() {
oldCluster := &apiv1.Cluster{
Spec: apiv1.ClusterSpec{
ImageName: versions.DefaultImageName,
ReplicationSlots: &apiv1.ReplicationSlotsConfiguration{
HighAvailability: &apiv1.ReplicationSlotsHAConfiguration{
Enabled: ptr.To(false),
},
SynchronizeReplicas: &apiv1.SynchronizeReplicasConfiguration{
Enabled: ptr.To(true),
},
},
},
}
Expand All @@ -3600,6 +3630,28 @@ var _ = Describe("validation of replication slots configuration", func() {
Expect(v.validateReplicationSlotsChange(newCluster, oldCluster)).To(HaveLen(1))
})

It("prevents removing the replication slot section when both features are enabled", func() {
oldCluster := &apiv1.Cluster{
Spec: apiv1.ClusterSpec{
ImageName: versions.DefaultImageName,
ReplicationSlots: &apiv1.ReplicationSlotsConfiguration{
HighAvailability: &apiv1.ReplicationSlotsHAConfiguration{
Enabled: ptr.To(true),
SlotPrefix: "_test_",
},
SynchronizeReplicas: &apiv1.SynchronizeReplicasConfiguration{
Enabled: ptr.To(true),
},
},
},
}
oldCluster.Default()

newCluster := oldCluster.DeepCopy()
newCluster.Spec.ReplicationSlots = nil
Expect(v.validateReplicationSlotsChange(newCluster, oldCluster)).To(HaveLen(2))
})

It("allows disabling the replication slots", func() {
oldCluster := &apiv1.Cluster{
Spec: apiv1.ClusterSpec{
Expand Down Expand Up @@ -3663,6 +3715,72 @@ var _ = Describe("validation of replication slots configuration", func() {
Expect(errors).To(BeEmpty())
})

It("prevents removing the synchronizeReplicas section when synchronizeReplicas is enabled", func() {
oldCluster := &apiv1.Cluster{
Spec: apiv1.ClusterSpec{
ImageName: versions.DefaultImageName,
ReplicationSlots: &apiv1.ReplicationSlotsConfiguration{
HighAvailability: &apiv1.ReplicationSlotsHAConfiguration{
Enabled: ptr.To(false),
},
SynchronizeReplicas: &apiv1.SynchronizeReplicasConfiguration{
Enabled: ptr.To(true),
},
},
},
}
oldCluster.Default()

newCluster := oldCluster.DeepCopy()
newCluster.Spec.ReplicationSlots.SynchronizeReplicas = nil
errs := v.validateReplicationSlotsChange(newCluster, oldCluster)
Expect(errs).To(HaveLen(1))
Expect(errs[0].Error()).To(ContainSubstring("Cannot remove"))
Expect(errs[0].Error()).To(ContainSubstring("synchronizeReplicas"))
})

It("allows disabling the synchronizeReplicas feature", func() {
oldCluster := &apiv1.Cluster{
Spec: apiv1.ClusterSpec{
ImageName: versions.DefaultImageName,
ReplicationSlots: &apiv1.ReplicationSlotsConfiguration{
HighAvailability: &apiv1.ReplicationSlotsHAConfiguration{
Enabled: ptr.To(false),
},
SynchronizeReplicas: &apiv1.SynchronizeReplicasConfiguration{
Enabled: ptr.To(true),
},
},
},
}
oldCluster.Default()

newCluster := oldCluster.DeepCopy()
newCluster.Spec.ReplicationSlots.SynchronizeReplicas.Enabled = ptr.To(false)
Expect(v.validateReplicationSlotsChange(newCluster, oldCluster)).To(BeEmpty())
})

It("allows removing synchronizeReplicas section when synchronizeReplicas is disabled", func() {
oldCluster := &apiv1.Cluster{
Comment thread
mnencia marked this conversation as resolved.
Resolved
Spec: apiv1.ClusterSpec{
ImageName: versions.DefaultImageName,
ReplicationSlots: &apiv1.ReplicationSlotsConfiguration{
HighAvailability: &apiv1.ReplicationSlotsHAConfiguration{
Enabled: ptr.To(false),
},
SynchronizeReplicas: &apiv1.SynchronizeReplicasConfiguration{
Enabled: ptr.To(false),
},
},
},
}
oldCluster.Default()

newCluster := oldCluster.DeepCopy()
newCluster.Spec.ReplicationSlots.SynchronizeReplicas = nil
Expect(v.validateReplicationSlotsChange(newCluster, oldCluster)).To(BeEmpty())
})

It("returns no errors when synchronizeLogicalDecoding is disabled", func() {
cluster := &apiv1.Cluster{
Spec: apiv1.ClusterSpec{
Expand Down
Loading