leonardBang commented on a change in pull request #15687:
URL: https://github.com/apache/flink/pull/15687#discussion_r619110800



##########
File path: docs/content.zh/release-notes/flink-1.13.md
##########
@@ -0,0 +1,227 @@
+---
+title: "Release Notes - Flink 1.13"
+---
+<!--
+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.
+-->
+
+# Release Notes - Flink 1.13
+
+These release notes discuss important aspects, such as configuration, 
behavior, or dependencies,
+that changed between Flink 1.12 and Flink 1.13. Please read these notes 
carefully if you are
+planning to upgrade your Flink version to 1.13.
+
+### Failover
+
+#### Remove state.backend.async option.
+
+##### [FLINK-21935](https://issues.apache.org/jira/browse/FLINK-21935)
+
+The `state.backend.async` option is deprecated. Snapshots are always 
asynchronous now (as they were
+by default before) and there is no option to configure a synchronous snapshot 
any more.
+
+The constructors of `FsStateBackend` and `MemoryStateBackend` that take a flag 
for sync/async
+snapshots are kept for API compatibility, but the flags are ignored now.
+
+#### Unify binary format for Keyed State savepoints
+
+##### [FLINK-20976](https://issues.apache.org/jira/browse/FLINK-20976)
+
+Flink’s savepoint binary format is unified across all state backends. That 
means you can take a
+savepoint with one state backend and then restore it using another.
+
+If you want to switch the state backend you should first upgrade your Flink 
version to 1.13, then
+take a savepoint with the new version, and only after that, you can restore it 
with a different
+state backend.
+
+#### FailureRateRestartBackoffTimeStrategy allows one less restart than 
configured
+
+##### [FLINK-20752](https://issues.apache.org/jira/browse/FLINK-20752)
+
+The Failure Rate Restart Strategy was allowing 1 less restart per interval 
than configured. Users
+wishing to keep the current behavior should reduce the maximum number of 
allowed failures per
+interval by 1.
+
+#### Support rescaling for Unaligned Checkpoints
+
+##### [FLINK-17979](https://issues.apache.org/jira/browse/FLINK-17979)
+
+While recovering from unaligned checkpoints, users can now change the 
parallelism of the job. This
+change allows users to quickly upscale the job under backpressure.
+
+### SQL
+
+#### Officially deprecate the legacy planner
+
+##### [FLINK-21709](https://issues.apache.org/jira/browse/FLINK-21709)
+
+The old planner of the Table & SQL API is deprecated and will be dropped in 
Flink 1.14. This means
+that both the BatchTableEnvironment and DataSet API interop are reaching end 
of life. Use the
+unified TableEnvironment for batch and stream processing with the new planner, 
or the DataStream API
+in batch execution mode.
+
+#### Disable problematic cast conversion between NUMERIC type and TIMESTAMP 
type
+
+##### [FLINK-21698](https://issues.apache.org/jira/browse/FLINK-21698)
+
+The CAST operation between `NUMERIC` type and `TIMESTAMP` type is problematic 
and is disabled now,
+e.g.
+`CAST(numeric AS TIMESTAMP(3))` is disabled and should use 
`TO_TIMESTAMP(FROM_UNIXTIME(numeric))`
+instead.

Review comment:
       ```suggestion
   #### Use TIMESTAMP_LTZ as return type for function PROCTIME()
   
   #### [FLINK-21714](https://issues.apache.org/jira/browse/FLINK-21714)
   
   Before Flink 1.13, the function return type of `PROCTIME()` is `TIMESTAMP`, 
and the return value is the `TIMESTAMP` in UTC time zone,
   e.g. the wall-clock shows `2021-03-01 12:00:00` at Shanghai, however the 
`PROCTIME()` displays `2021-03-01 04:00:00` which is wrong.
   Flink 1.13 fixes this issue and uses `TIMESTAMP_LTZ` type as return type of 
`PROCTIME()`, users don't need to deal time zone problems anymore.
   
   #### Support defining event time attribute on TIMESTAMP_LTZ column
   
   #### [FLINK-20387](https://issues.apache.org/jira/browse/FLINK-20387)
   
   Support defining event time attribute on TIMESTAMP_LTZ column, base on this, 
Flink SQL gracefully support the Daylight Saving Time. 
   
   #### Correct function 
CURRENT_TIMESTAMP/CURRENT_TIME/CURRENT_DATE/LOCALTIME/LOCALTIMESTAMP/NOW()
   
   ##### [FLINK-21713](https://issues.apache.org/jira/browse/FLINK-21713)
   
   The value of time function CURRENT_TIMESTAMP and NOW() are corrected from 
UTC time with `TIMESTAMP` type to epoch time with `TIMESTAMP_LTZ` type. Time 
function LOCALTIME, LOCALTIMESTAMP, CURRENT_DATE, CURRENT_TIME, 
CURRENT_TIMESTAMP and NOW() are corrected from evaluates for per record in 
batch mode to evaluate once at query-start for batch job.
   
   #### Disable problematic cast conversion between NUMERIC type and TIMESTAMP 
type
   
   ##### [FLINK-21698](https://issues.apache.org/jira/browse/FLINK-21698)
   
   The CAST operation between `NUMERIC` type and `TIMESTAMP` type is 
problematic and is disabled now,
   e.g.
   `CAST(numeric AS TIMESTAMP(3))` is disabled and should use 
`TO_TIMESTAMP(FROM_UNIXTIME(numeric))`
   instead.
   ```
   
   would be better if we can add these incompatibility note here, @dawidwys  




-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
[email protected]


Reply via email to