From a205a838298841d780056d9c21c4340076214d99 Mon Sep 17 00:00:00 2001
From: Varun chandu Pakalapati <pakalapa@Varuns-MacBook-Pro.local>
Date: Sun, 14 Dec 2025 08:53:09 -0600
Subject: [PATCH] fix: prevent reconciliation loop for parameters pending
 reboot

- Skip modifying/resetting parameters that are already pending-reboot
- Prevents continuous loop when parameters are in transitional state
- Applies to both DBClusterParameterGroup and DBParameterGroup
- Fixes issue where ACK continuously tries to modify parameters after reboot

The loop occurred because:
1. Parameter set to pending-reboot status
2. After reboot, parameter at default but status still shows pending-reboot
3. ACK sees mismatch and tries to modify/reset again
4. Loop continues indefinitely

Now ACK checks ParameterOverrideStatuses before modifying/resetting
parameters and skips any that are already pending-reboot.
---
 .../db_cluster_parameter_group/hooks.go       | 53 +++++++++++++++++++
 pkg/resource/db_parameter_group/hooks.go      | 53 +++++++++++++++++++
 2 files changed, 106 insertions(+)

diff --git a/pkg/resource/db_cluster_parameter_group/hooks.go b/pkg/resource/db_cluster_parameter_group/hooks.go
index a3478bc8..01df363c 100644
--- a/pkg/resource/db_cluster_parameter_group/hooks.go
+++ b/pkg/resource/db_cluster_parameter_group/hooks.go
@@ -242,6 +242,59 @@ func (rm *resourceManager) syncParameters(
 		desiredOverrides, latestOverrides,
 	)
 
+	// Filter out parameters that are already pending reboot from both toModify and toDelete.
+	// When a parameter is reset or modified, it remains as a user override with pending-reboot
+	// status until the DB cluster is rebooted. Attempting to modify or reset it again
+	// would cause a reconciliation loop. We skip modifying/resetting parameters that are
+	// already pending reboot until after the reboot completes.
+	if latest != nil && latest.ko.Status.ParameterOverrideStatuses != nil {
+		pendingRebootParams := make(map[string]bool)
+		
+		// Build a map of parameters that are pending reboot
+		for _, status := range latest.ko.Status.ParameterOverrideStatuses {
+			if status.ParameterName != nil && status.ApplyMethod != nil {
+				if *status.ApplyMethod == "pending-reboot" {
+					pendingRebootParams[*status.ParameterName] = true
+				}
+			}
+		}
+		
+		// Filter toDelete: skip parameters that are pending reboot
+		if len(toDelete) > 0 {
+			filteredToDelete := util.Parameters{}
+			for paramName, paramValue := range toDelete {
+				if !pendingRebootParams[paramName] {
+					filteredToDelete[paramName] = paramValue
+				} else {
+					rlog.Debug(
+						"skipping reset of parameter already pending reboot",
+						"parameter", paramName,
+					)
+				}
+			}
+			toDelete = filteredToDelete
+		}
+		
+		// Filter toModify: skip parameters that are pending reboot
+		// This prevents trying to modify parameters that are in a transitional state
+		// after reboot but before the status is updated
+		if len(toModify) > 0 {
+			filteredToModify := util.Parameters{}
+			for paramName, paramValue := range toModify {
+				if !pendingRebootParams[paramName] {
+					filteredToModify[paramName] = paramValue
+				} else {
+					rlog.Debug(
+						"skipping modify of parameter already pending reboot",
+						"parameter", paramName,
+						"desired_value", paramValue,
+					)
+				}
+			}
+			toModify = filteredToModify
+		}
+	}
+
 	if len(toDelete) > 0 {
 		err = rm.resetParameters(ctx, family, groupName, toDelete)
 		if err != nil {
diff --git a/pkg/resource/db_parameter_group/hooks.go b/pkg/resource/db_parameter_group/hooks.go
index 41b41d80..c496bbd8 100644
--- a/pkg/resource/db_parameter_group/hooks.go
+++ b/pkg/resource/db_parameter_group/hooks.go
@@ -230,6 +230,59 @@ func (rm *resourceManager) syncParameters(
 		desiredOverrides, latestOverrides,
 	)
 
+	// Filter out parameters that are already pending reboot from both toModify and toDelete.
+	// When a parameter is reset or modified, it remains as a user override with pending-reboot
+	// status until the DB instance is rebooted. Attempting to modify or reset it again
+	// would cause a reconciliation loop. We skip modifying/resetting parameters that are
+	// already pending reboot until after the reboot completes.
+	if latest != nil && latest.ko.Status.ParameterOverrideStatuses != nil {
+		pendingRebootParams := make(map[string]bool)
+		
+		// Build a map of parameters that are pending reboot
+		for _, status := range latest.ko.Status.ParameterOverrideStatuses {
+			if status.ParameterName != nil && status.ApplyMethod != nil {
+				if *status.ApplyMethod == "pending-reboot" {
+					pendingRebootParams[*status.ParameterName] = true
+				}
+			}
+		}
+		
+		// Filter toDelete: skip parameters that are pending reboot
+		if len(toDelete) > 0 {
+			filteredToDelete := util.Parameters{}
+			for paramName, paramValue := range toDelete {
+				if !pendingRebootParams[paramName] {
+					filteredToDelete[paramName] = paramValue
+				} else {
+					rlog.Debug(
+						"skipping reset of parameter already pending reboot",
+						"parameter", paramName,
+					)
+				}
+			}
+			toDelete = filteredToDelete
+		}
+		
+		// Filter toModify: skip parameters that are pending reboot
+		// This prevents trying to modify parameters that are in a transitional state
+		// after reboot but before the status is updated
+		if len(toModify) > 0 {
+			filteredToModify := util.Parameters{}
+			for paramName, paramValue := range toModify {
+				if !pendingRebootParams[paramName] {
+					filteredToModify[paramName] = paramValue
+				} else {
+					rlog.Debug(
+						"skipping modify of parameter already pending reboot",
+						"parameter", paramName,
+						"desired_value", paramValue,
+					)
+				}
+			}
+			toModify = filteredToModify
+		}
+	}
+
 	// NOTE(jaypipes): ResetDBParameterGroup and ModifyDBParameterGroup only
 	// accept 20 parameters at a time, which is why we "chunk" both the deleted
 	// and modified parameter sets.