[
https://issues.apache.org/jira/browse/FLINK-7301?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16112486#comment-16112486
]
ASF GitHub Bot commented on FLINK-7301:
---------------------------------------
Github user alpinegizmo commented on a diff in the pull request:
https://github.com/apache/flink/pull/4441#discussion_r131086039
--- Diff: docs/dev/stream/state/custom_serialization.md ---
@@ -0,0 +1,188 @@
+---
+title: "Custom Serialization for Managed State"
+nav-title: "Custom Serialization"
+nav-parent_id: streaming_state
+nav-pos: 10
+---
+<!--
+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.
+-->
+
+If your application uses Flink's managed state, it might be necessary to
implement a custom serialization logic for special use cases.
+
+This page is targeted as a guideline for users who require the use of
custom serialization for their state, covering how
+to provide a custom serializer and how to handle upgrades to the
serializer for compatibility. If you're simply using
+Flink's own serializers, this page is irrelevant and can be skipped.
+
+### Using custom serializers
+
+As demonstrated in the above examples, when registering a managed operator
or keyed state, a `StateDescriptor` is required
+to specify the state's name, as well as information about the type of the
state. The type information is used by Flink's
+[type serialization framework](../../types_serialization.html) to create
appropriate serializers for the state.
+
+It is also possible to completely bypass this and let Flink use your own
custom serializer to serialize managed states,
+simply by directly instantiating the `StateDescriptor` with your own
`TypeSerializer` implementation:
+
+<div class="codetabs" markdown="1">
+<div data-lang="java" markdown="1">
+{% highlight java %}
+public class CustomTypeSerializer extends TypeSerializer<Tuple2<String,
Integer>> {...};
+
+ListStateDescriptor<Tuple2<String, Integer>> descriptor =
+ new ListStateDescriptor<>(
+ "state-name",
+ new CustomTypeSerializer());
+
+checkpointedState = getRuntimeContext().getListState(descriptor);
+{% endhighlight %}
+</div>
+
+<div data-lang="scala" markdown="1">
+{% highlight scala %}
+class CustomTypeSerializer extends TypeSerializer[(String, Integer)] {...}
+
+val descriptor = new ListStateDescriptor[(String, Integer)](
+ "state-name",
+ new CustomTypeSerializer)
+)
+
+checkpointedState = getRuntimeContext.getListState(descriptor);
+{% endhighlight %}
+</div>
+</div>
+
+Note that Flink writes state serializers along with the state as metadata.
In certain cases on restore (see following
+subsections), the written serializer needs to be deserialized and used.
Therefore, it is recommended to avoid using
+anonymous classes as your state serializers. Anonymous classes do not have
a guarantee on the generated classname,
+varying across compilers and depends on the order that they are
instantiated within the enclosing class, which can
--- End diff --
"varying across compilers and depends" ==> "which varies across compilers
and depends"
> Rework state documentation
> --------------------------
>
> Key: FLINK-7301
> URL: https://issues.apache.org/jira/browse/FLINK-7301
> Project: Flink
> Issue Type: Improvement
> Components: Documentation
> Reporter: Timo Walther
> Assignee: Timo Walther
>
> The documentation about state is spread across different pages, but this is
> not consistent and it is hard to find what you need. I propose:
> "Mention State Backends and link to them in ""Streaming/Working with State"".
> Create category ""State & Fault Tolerance"" under ""Streaming"". Move
> ""Working with State"", ""Checkpointing"" and ""Queryable State"".
> Move API related parts (90%) of ""Deployment/State & Fault Tolerance/State
> Backends"" to ""Streaming/State & Fault Tolerance/State Backends"".
> Move all tuning things from ""Debugging/Large State"" to ""Deployment/State &
> Fault Tolerance/State Backends"".
> Move ""Streaming/Working with State/Custom Serialization for Managed State""
> to ""Streaming/State & Fault Tolerance/Custom Serialization"" (Add a link
> from previous position, also link from ""Data Types & Serialization"")."
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
