From b1e5454ad7c4b0b8bbae93d509bd1c05d8477c1b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sinclert=20P=C3=A9rez?= <sinclert.perez@canonical.com>
Date: Mon, 20 Jan 2025 12:16:05 +0100
Subject: [PATCH] Add cluster manual rejoin mechanism

---
 src/charm.py | 33 ++++++++++++++++++++++++++++++++-
 1 file changed, 32 insertions(+), 1 deletion(-)

diff --git a/src/charm.py b/src/charm.py
index 91c2ef7a3..4e6806bf2 100755
--- a/src/charm.py
+++ b/src/charm.py
@@ -449,7 +449,7 @@ def _handle_non_online_instance_status(self, state) -> None:
             }
             all_states.add("offline")
 
-            if all_states == {"offline"} and self.unit.is_leader():
+            if self.unit.is_leader() and all_states == {"offline"}:
                 # All instance are off or its a single unit cluster
                 # reboot cluster from outage from the leader unit
                 logger.info("Attempting reboot from complete outage.")
@@ -460,6 +460,13 @@ def _handle_non_online_instance_status(self, state) -> None:
                     logger.error("Failed to reboot cluster from complete outage.")
                     self.unit.status = BlockedStatus("failed to recover cluster.")
 
+            elif self._mysql.is_cluster_auto_rejoin_ongoing():
+                logger.info("Cluster auto-rejoin attempts are still ongoing.")
+
+            elif not self._mysql.is_cluster_auto_rejoin_ongoing():
+                logger.info("Cluster auto-rejoin attempts are exhausted. Attempting manual rejoin")
+                self._execute_manual_rejoin()
+
         if state == "unreachable":
             try:
                 if not snap_service_operation(
@@ -471,6 +478,30 @@ def _handle_non_online_instance_status(self, state) -> None:
             except SnapServiceOperationError as e:
                 self.unit.status = BlockedStatus(e.message)
 
+    def _execute_manual_rejoin(self) -> None:
+        """Executes an instance manual rejoin.
+
+        It is supposed to be called when the MySQL 8.0.21+ auto-rejoin attempts have been exhausted,
+        on an OFFLINE replica that still belongs to the cluster
+        """
+        if not self._mysql.is_instance_in_cluster(self.unit_label):
+            logger.warning("Instance does not belong to the cluster. Cannot perform manual rejoin")
+            return
+
+        cluster_primary = self._get_primary_from_online_peer()
+        if not cluster_primary:
+            logger.warning("Instance does not have ONLINE peers. Cannot perform manual rejoin")
+            return
+
+        self._mysql.remove_instance(
+            unit_label=self.unit_label,
+        )
+        self._mysql.add_instance_to_cluster(
+            instance_address=self.unit_address,
+            instance_unit_label=self.unit_label,
+            from_instance=cluster_primary,
+        )
+
     def _on_update_status(self, _) -> None:  # noqa: C901
         """Handle update status.