Bug 37393929 - [37181866->24.09.1] RFA: Snapshot creation failed disa…

…bling active persistence with IllegalArgumentException: unknown extent identifier [git-p4: depot-paths = "//dev/coherence-ce/release/coherence-ce-v24.09/": change = 113584]
oracle · Jan 16, 2025 · bf6e4c8 · bf6e4c8
1 parent 42e4329
commit bf6e4c8
Show file tree

Hide file tree

Showing 3 changed files with 42 additions and 14 deletions.
diff --git a/...gosol/coherence/component/util/daemon/queueProcessor/service/grid/PartitionedService.java b/...gosol/coherence/component/util/daemon/queueProcessor/service/grid/PartitionedService.java
@@ -2292,7 +2292,8 @@ public boolean enterPartition(int nPartition, long cWait)
         int nLockType = ctrlPartition.getLockType();
         if (nLockType == PartitionedService.PartitionControl.LOCK_BACKUP_XFER_OUT ||
             nLockType == PartitionedService.PartitionControl.LOCK_NONE ||
-            nLockType == PartitionedService.PartitionControl.LOCK_PERSISTENCE)
+            nLockType == PartitionedService.PartitionControl.LOCK_PERSISTENCE ||
+            nLockType == PartitionedService.PartitionControl.LOCK_PERSISTENCE_SNAPSHOT)
             {
             ctrlPartition.enter(-1L);
             return true;
@@ -20996,13 +20997,20 @@ public static abstract class PartitionControl
          *
          * Lock is pending.
          */
-        public static final int LOCK_PENDING = 8;
-        
+        public static final int LOCK_PENDING = 16;
+
         /**
-         * Property LOCK_PERSISTENCE
+         * Property LOCK_PERSISTENCE_SNAPSHOT
          *
          * Lock for partition persistence snapshot.
          */
+        public static final int LOCK_PERSISTENCE_SNAPSHOT = 8;
+
+        /**
+         * Property LOCK_PERSISTENCE
+         *
+         * Lock for partition persistence recovery.
+         */
         public static final int LOCK_PERSISTENCE = 4;
 
         /**
@@ -21162,7 +21170,7 @@ public static abstract class PartitionControl
          * 
          * 0x0F (bits 1-3)
          */
-        public static final int STATE_MASK_LOCK = 15;
+        public static final int STATE_MASK_LOCK = 31;
 
         /**
          * Property STATE_MASK_PIN
@@ -27743,7 +27751,7 @@ protected boolean lock(int iPart)
                 PartitionedService.PartitionControl ctrlPart = service.getPartitionControl(iPart);
                 long              cMillis  = service.getDistributionContendMillis();
                 boolean           fLocked  = ctrlPart != null && // a once owned partition may no longer be owned
-                                             ctrlPart.lock(cMillis, PartitionedService.PartitionControl.LOCK_PERSISTENCE);
+                                             ctrlPart.lock(cMillis, PartitionedService.PartitionControl.LOCK_PERSISTENCE_SNAPSHOT);
 
                 AtomicInteger atomicTasks = ctrlPart.getPersistenceTasks();
                 if (fLocked && atomicTasks.get() > 0)
@@ -28090,7 +28098,7 @@ protected void snapshotStore(int iPart)
                     else
                         {
                         PersistentStore storeFrom = service.getPartitionControl(iPart).getPersistentStore();
-                
+
                         store = mgrSnapshot.open(sGUID = storeFrom.getId(), storeFrom);
 
                         removeTransientCaches(store);

diff --git a/.../util/daemon/queueProcessor/service/grid/partitionedService/partitionedCache/Storage.java b/.../util/daemon/queueProcessor/service/grid/partitionedService/partitionedCache/Storage.java
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2000, 2024, Oracle and/or its affiliates.
+ * Copyright (c) 2000, 2025, Oracle and/or its affiliates.
  *
  * Licensed under the Universal Permissive License v 1.0 as shown at
  * https://oss.oracle.com/licenses/upl.
@@ -6591,6 +6591,7 @@ public void onBackingMapEvent(com.tangosol.util.MapEvent evt)
                         // state is changing; try again.
                         break;
 
+                    case PartitionedCache.PartitionControl.LOCK_PERSISTENCE_SNAPSHOT:
                     case PartitionedCache.PartitionControl.LOCK_PERSISTENCE:
                     case PartitionedCache.PartitionControl.LOCK_PRIMARY_XFER_IN:
                     case PartitionedCache.PartitionControl.LOCK_BACKUP_XFER_OUT:

diff --git a/prj/coherence-core/src/main/java/com/tangosol/persistence/AbstractPersistenceManager.java b/prj/coherence-core/src/main/java/com/tangosol/persistence/AbstractPersistenceManager.java
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2000, 2024, Oracle and/or its affiliates.
+ * Copyright (c) 2000, 2025, Oracle and/or its affiliates.
  *
  * Licensed under the Universal Permissive License v 1.0 as shown at
  * https://oss.oracle.com/licenses/upl.
@@ -274,7 +274,7 @@ public PersistentStoreInfo[] listStoreInfo()
             return new PersistentStoreInfo[0];
             }
 
-        List<PersistentStoreInfo> listInfo = new ArrayList<PersistentStoreInfo>();
+        List<PersistentStoreInfo> listInfo = new ArrayList<>();
 
         for (File fileEnv : aFiles)
             {
@@ -1375,7 +1375,12 @@ public void store(long lExtentId, ReadBuffer bufKey, ReadBuffer bufValue, Object
             lockRead();
             try
                 {
-                validateExtentId(lExtentId);
+                if (!validateExtentId(lExtentId))
+                    {
+                    // ignore normal cases such as concurrent cache deletes, abnormal ones
+                    // are handled using an exception.
+                    return;
+                    }
 
                 if (oToken instanceof AbstractPersistenceManager.AbstractPersistentStore.BatchTask)
                     {
@@ -1435,7 +1440,12 @@ public void erase(long lExtentId, ReadBuffer bufKey, Object oToken)
             lockRead();
             try
                 {
-                validateExtentId(lExtentId);
+                if (!validateExtentId(lExtentId))
+                    {
+                    // ignore normal cases such as concurrent cache deletes, abnormal ones
+                    // are handled using an exception.
+                    return;
+                    }
 
                 if (oToken instanceof AbstractPersistenceManager.AbstractPersistentStore.BatchTask)
                     {
@@ -2048,15 +2058,24 @@ protected final void unlockWrite()
          *
          * @param lExtentId  the extent identifier
          */
-        protected void validateExtentId(long lExtentId)
+        protected boolean validateExtentId(long lExtentId)
             {
             Long LId = Long.valueOf(lExtentId);
 
             // validate that the given extent identifier is known
             if (!f_setExtentIds.contains(LId))
                 {
-                throw new IllegalArgumentException("unknown extent identifier: " + lExtentId + " for store: " + getId());
+                if (f_setDeletedIds.contains(LId))
+                    {
+                    Logger.warn("extent identifier " + LId + " has been concurrently deleted for store: " + getId());
+                    return false;
+                    }
+                else
+                    {
+                    throw new IllegalArgumentException("unknown extent identifier: " + lExtentId + " for store: " + getId());
+                    }
                 }
+            return true;
             }
 
         /**