flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (FLINK-8360) Implement task-local state recovery
Date Fri, 16 Feb 2018 14:42:00 GMT

    [ https://issues.apache.org/jira/browse/FLINK-8360?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16367389#comment-16367389
] 

ASF GitHub Bot commented on FLINK-8360:
---------------------------------------

Github user StefanRRichter commented on a diff in the pull request:

    https://github.com/apache/flink/pull/5239#discussion_r168773814
  
    --- Diff: flink-contrib/flink-statebackend-rocksdb/src/main/java/org/apache/flink/contrib/streaming/state/RocksDBKeyedStateBackend.java
---
    @@ -266,11 +305,16 @@ public RocksDBKeyedStateBackend(
     		RocksIterator iterator = db.newIterator(columnInfo.f0);
     		iterator.seekToFirst();
     
    -		Iterable<K> iterable = () -> new RocksIteratorWrapper<>(iterator, state,
keySerializer, keyGroupPrefixBytes);
    +		Iterable<K> iterable = () -> new RocksIteratorToJavaIteratorAdapter<>(iterator,
state, keySerializer, keyGroupPrefixBytes);
     		Stream<K> targetStream = StreamSupport.stream(iterable.spliterator(), false);
    --- End diff --
    
    Fine with me, but since this seems to have not caused any trouble so far, you might first
want to check if the namespace parameter is actually required. Maybe it can just be removed
as parameter because the using code wants to iterate all keys in all namespace.


> Implement task-local state recovery
> -----------------------------------
>
>                 Key: FLINK-8360
>                 URL: https://issues.apache.org/jira/browse/FLINK-8360
>             Project: Flink
>          Issue Type: New Feature
>          Components: State Backends, Checkpointing
>            Reporter: Stefan Richter
>            Assignee: Stefan Richter
>            Priority: Major
>             Fix For: 1.5.0
>
>
> This issue tracks the development of recovery from task-local state. The main idea is
to have a secondary, local copy of the checkpointed state, while there is still a primary
copy in DFS that we report to the checkpoint coordinator.
> Recovery can attempt to restore from the secondary local copy, if available, to save
network bandwidth. This requires that the assignment from tasks to slots is as sticky is possible.
> For starters, we will implement this feature for all managed keyed states and can easily
enhance it to all other state types (e.g. operator state) later.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message