Github user StefanRRichter commented on a diff in the pull request:
https://github.com/apache/flink/pull/3558#discussion_r108133263
--- Diff:
flink-runtime/src/main/java/org/apache/flink/migration/v0/SavepointV0.java ---
@@ -0,0 +1,207 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.flink.migration.v0;
+
+import org.apache.flink.migration.v0.runtime.TaskStateV0;
+import org.apache.flink.runtime.checkpoint.TaskState;
+import org.apache.flink.runtime.checkpoint.savepoint.Savepoint;
+import org.apache.flink.util.Preconditions;
+
+import java.util.Collection;
+import java.util.HashMap;
+import java.util.Map;
+
+/**
+ * Savepoint version 0.
+ *
+ * <p>This format was introduced with Flink 1.1.0.
+ * <pre>
+ * checkpointId: long
+ * numTaskStates: int
+ * |----jobVertexID: long[2]
+ * |----parallelism: int
+ * |----numSubtaskStates: int
+ * | |----subtaskIndex: int
+ * | |----serializedValueLength: int
+ * | |----serializedValue: byte[] (null if serializedValueLength is
-1)
+ * | |----subtaskStateSize: long
+ * | |----subtaskStateDuration: long
+ * |----numKeyGroupStates: int
+ * | |----subtaskIndex: int
+ * | |----serializedValueLength: int
+ * | |----serializedValue: byte[] (null if serializedValueLength is
-1)
+ * | |----keyGroupStateSize: long
+ * | |----keyGroupStateDuration: long
+ * </pre>
+ */
+@Deprecated
+@SuppressWarnings("deprecation")
+public class SavepointV0 implements Savepoint {
+
+ /** The classes that are migrated in SavepointV0 */
+ public static final Map<String, String> MigrationMapping = new
HashMap<String, String>() {{
+
+ /* migrated state descriptors */
+ put("org.apache.flink.api.common.state.StateDescriptor",
--- End diff --
I wonder if it is a good idea to make the mapping as explicit as this.
While this offers more flexibility than the previous approach, we also need to
maintain this mapping by hand. This can be error prone, in case a class is
forgotten or there is a typo. Would it make sense to either a) create the
string mappings in this map automatically through code that explores classes in
the migration package or b) keep a naming scheme convention, similar to before
but with the version attached in the package in class name? Then we could use
this map to only contain mappings that we want to be explicitly different from
to convention (is there a case for this?). My feeling is that option b) might
be better. What do you think?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
