[REPO-5592] Support for PostgreSQL and MySQL, batch size is now configurable (#485)

* Initial support for PosgreSQL, MySQL plus batchSize is now configurable

* Changes post review

repository/src/main/java/org/alfresco/repo/content/cleanup/ContentStoreCleaner.java

@@ -2,7 +2,7 @@
  * #%L
  * Alfresco Repository
  * %%
- * Copyright (C) 2005 - 2016 Alfresco Software Limited
+ * Copyright (C) 2005 - 2021 Alfresco Software Limited
  * %%
  * This file is part of the Alfresco software. 
  * If the software was purchased under a paid Alfresco license, the terms of 
@@ -125,10 +125,12 @@ public class ContentStoreCleaner
     private ContentService contentService;
     private TransactionService transactionService;
     private int protectDays;
+    private int batchSize;
     private DeleteFailureAction deletionFailureAction;
     
     public ContentStoreCleaner()
     {
+        this.batchSize = 1000;
         this.protectDays = 7;
         this.deletionFailureAction = DeleteFailureAction.IGNORE;
     }
@@ -192,6 +194,16 @@ public class ContentStoreCleaner
         this.protectDays = protectDays;
     }
 
+    /**
+     * Set the batch size used by the cleaning jobs. The default is 1000
+     * 
+     * @param batchSize     batch size for each cleanup job
+     */
+    public void setBatchSize(int batchSize)
+    {
+        this.batchSize = batchSize;
+    }
+
     /**
      * Set the action to take in the event that an orphaned binary failed to get deleted.
      * The default is {@link DeleteFailureAction#IGNORE}.
@@ -329,7 +341,7 @@ public class ContentStoreCleaner
         {
             public Long execute() throws Exception
             {
-                return cleanBatch(maxOrphanTime, 1000);
+                return cleanBatch(maxOrphanTime, batchSize);
             };
         };
         while (true)

repository/src/main/java/org/alfresco/repo/domain/contentdata/ContentUrlOrphanQuery.java

@@ -2,7 +2,7 @@
  * #%L
  * Alfresco Repository
  * %%
- * Copyright (C) 2005 - 2016 Alfresco Software Limited
+ * Copyright (C) 2005 - 2021 Alfresco Software Limited
  * %%
  * This file is part of the Alfresco software. 
  * If the software was purchased under a paid Alfresco license, the terms of 
@@ -34,6 +34,7 @@ package org.alfresco.repo.domain.contentdata;
 public class ContentUrlOrphanQuery
 {
     private Long maxOrphanTimeExclusive;
+    private Long maxRecords;
     
     @Override
     public String toString()
@@ -41,6 +42,7 @@ public class ContentUrlOrphanQuery
         StringBuilder sb = new StringBuilder(512);
         sb.append("ContentUrlOrphanQuery")
           .append("[ maxOrphanTimeExclusive=").append(maxOrphanTimeExclusive)
+          .append(", maxRecords=").append(maxRecords)
           .append("]");
         return sb.toString();
     }
@@ -53,5 +55,17 @@ public class ContentUrlOrphanQuery
     public void setMaxOrphanTimeExclusive(Long maxOrphanTimeExclusive)
     {
         this.maxOrphanTimeExclusive = maxOrphanTimeExclusive;
-    }
+    }
+
+    public Long getMaxRecords()
+    {
+        return maxRecords;
+    }
+
+    public void setMaxRecords(Long maxRecords)
+    {
+        this.maxRecords = maxRecords;
+    }
+    
+    
 }

repository/src/main/java/org/alfresco/repo/domain/contentdata/ibatis/ContentDataDAOImpl.java

@@ -2,7 +2,7 @@
  * #%L
  * Alfresco Repository
  * %%
- * Copyright (C) 2005 - 2016 Alfresco Software Limited
+ * Copyright (C) 2005 - 2021 Alfresco Software Limited
  * %%
  * This file is part of the Alfresco software. 
  * If the software was purchased under a paid Alfresco license, the terms of 
@@ -155,6 +155,7 @@ public class ContentDataDAOImpl extends AbstractContentDataDAOImpl
         
         ContentUrlOrphanQuery query = new ContentUrlOrphanQuery();
         query.setMaxOrphanTimeExclusive(maxOrphanTimeExclusive);
+        query.setMaxRecords((long) maxResults);
         List<ContentUrlEntity> results = template.selectList(SELECT_CONTENT_URLS_ORPHANED, 
                                                                                       query, 
                                                                                       new RowBounds(0, maxResults));

repository/src/main/resources/alfresco/content-services-context.xml

@@ -115,6 +115,9 @@
       <property name="transactionService" >
          <ref bean="transactionService" />
       </property>
+      <property name="batchSize" >
+         <value>${system.content.cleanerBatchSize}</value>
+      </property>
    </bean>
 
    <bean id="eagerContentStoreCleaner" class="org.alfresco.repo.content.cleanup.EagerContentStoreCleaner" init-method="init">

repository/src/main/resources/alfresco/ibatis/org.alfresco.repo.domain.dialect.Dialect/content-select-SqlMap.xml

@@ -18,9 +18,10 @@
             left outer join alf_content_data cd on (cd.content_url_id = cu.id)
         where
             cd.id is null and
+            cu.orphan_time is not null and
             cu.orphan_time > 0 and
-            cu.orphan_time < #{maxOrphanTimeExclusive} and
-            cu.orphan_time is not null
+            cu.orphan_time < #{maxOrphanTimeExclusive}
+        limit #{maxRecords}
         ]]>
     </select>
 

repository/src/main/resources/alfresco/ibatis/org.alfresco.repo.domain.dialect.MySQLInnoDBDialect/content-select-SqlMap.xml

@@ -21,6 +21,7 @@
             cu.orphan_time > 0 and
             cu.orphan_time < #{maxOrphanTimeExclusive} and
             cu.orphan_time is not null
+        limit #{maxRecords}
         ]]>
     </select>
 

repository/src/main/resources/alfresco/repository.properties

@@ -219,6 +219,8 @@ system.content.orphanProtectDays=14
 system.content.deletionFailureAction=IGNORE
 # The CRON expression to trigger the deletion of resources associated with orphaned content.
 system.content.orphanCleanup.cronExpression=0 0 4 * * ?
+# The batch size user by the content store cleaner
+system.content.cleanerBatchSize=1000
 
 # The CRON expression to trigger the cleanup of deleted nodes and dangling transactions that are old enough
 system.nodeServiceCleanup.cronExpression=0 0 21 * * ?