[
https://issues.apache.org/jira/browse/FLINK-7213?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16126150#comment-16126150
]
ASF GitHub Bot commented on FLINK-7213:
---------------------------------------
Github user StephanEwen commented on a diff in the pull request:
https://github.com/apache/flink/pull/4353#discussion_r133010887
--- Diff:
flink-runtime/src/main/java/org/apache/flink/runtime/checkpoint/TaskStateSnapshot.java
---
@@ -0,0 +1,139 @@
+/*
+ * 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.runtime.checkpoint;
+
+import org.apache.flink.runtime.jobgraph.OperatorID;
+import org.apache.flink.runtime.state.CompositeStateHandle;
+import org.apache.flink.runtime.state.SharedStateRegistry;
+import org.apache.flink.runtime.state.StateUtil;
+import org.apache.flink.util.Preconditions;
+
+import java.util.HashMap;
+import java.util.Map;
+import java.util.Set;
+
+/**
+ * This class encapsulates state handles to the snapshots of all operator
instances executed within one task. A task
+ * can run multiple operator instances as a result of operator chaining,
and all operator instances from the chain can
+ * register their state under their operator id. Each operator instance is
a physical execution responsible for
+ * processing a partition of the data that goes through a logical
operator. This partitioning happens to parallelize
+ * execution of logical operators, e.g. distributing a map function.
+ *
+ * <p>One instance of this class contains the information that one task
will send to acknowledge a checkpoint request by
+ * the checkpoint coordinator. Tasks run operator instances in parallel,
so the union of all
+ * {@link TaskStateSnapshot} that are collected by the checkpoint
coordinator from all tasks represent the whole
+ * state of a job at the time of the checkpoint.
+ *
+ * <p>This class should be called TaskState once the old class with this
name that we keep for backwards
+ * compatibility goes away.
+ */
+public class TaskStateSnapshot implements CompositeStateHandle {
--- End diff --
Would it make sense to make this immutable? It looks like this should not
be modified any more after fully constructing it. This would also make it clear
that methods iterating over the state, or returning sets / iterables can never
fail with concurrent modifications.
For example the `size` method is considered a "best effort" method for info
purposes only, and should not fail with an exception (it currently could fail
with a `ConcurrentModificationException`).
> Introduce state management by OperatorID in TaskManager
> -------------------------------------------------------
>
> Key: FLINK-7213
> URL: https://issues.apache.org/jira/browse/FLINK-7213
> Project: Flink
> Issue Type: Improvement
> Components: State Backends, Checkpointing
> Affects Versions: 1.4.0
> Reporter: Stefan Richter
> Assignee: Stefan Richter
>
> Flink-5892 introduced the job manager / checkpoint coordinator part of
> managing state on the operator level instead of the task level by introducing
> explicit operator_id -> state mappings. However, this explicit mapping was
> not introduced in the task manager side, so the explicit mapping is still
> converted into a mapping that suits the implicit operator chain order.
> We should also introduce explicit operator ids to state management on the
> task manager.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
