[ 
https://issues.apache.org/jira/browse/FLINK-12406?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Ruidong Li updated FLINK-12406:
-------------------------------
    Description: 
After each job finishes, the new {{BLOCKING_PERSISTENT}} result partitions are 
generated, and locations of these result partitions should be report back to 
client via {{JobExecutionResult}}, they will be later used for Table 
{{cache()}} and {{invalidateCache()}}

 

Brief Changes:

- Add a new class {{IntermediateResultDescriptor}}, which stores location of a 
{{BLOCKING_PERSISTENT}} {{ResultPartition}}, currently we only support 
{{ResultPartition}} in {{TaskManager}}.
 - Add a new method {{getResultPartitionDescriptors()}} in 
{{AccessExecutionGraph}}
 - Add a new filed in {{JobExecutionResult}}, {{JobResult}}, 
{{ArchivedExecutionGraph}} and {{ExecutionEnvironment}}, which keeps a mapping 
from {{IntermediateDataSetID}} to its {{ResultPartition}} locations
 - When a job finishes, the metadata will flow in this path: {{ExecutionGraph}} 
-> {{ArchivedExecutionGraph}} -> {{JobExecutionResult}} -> {{JobResult}} -> 
{{ExecutionEnvironment}}

  was:After each job finishes, the new {{BLOCKING_PERSISTENT}} result 
partitions are generated, and locations of these result partitions should be 
report back to client via {{JobExecutionResult}}, they will be later used for 
Table {{cache()}} and {{invalidateCache()}}


> Report BLOCKING_PERSISTENT result partition meta back to client
> ---------------------------------------------------------------
>
>                 Key: FLINK-12406
>                 URL: https://issues.apache.org/jira/browse/FLINK-12406
>             Project: Flink
>          Issue Type: Sub-task
>          Components: API / DataSet, Runtime / Coordination
>            Reporter: Ruidong Li
>            Assignee: Ruidong Li
>            Priority: Major
>              Labels: pull-request-available
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> After each job finishes, the new {{BLOCKING_PERSISTENT}} result partitions 
> are generated, and locations of these result partitions should be report back 
> to client via {{JobExecutionResult}}, they will be later used for Table 
> {{cache()}} and {{invalidateCache()}}
>  
> Brief Changes:
> - Add a new class {{IntermediateResultDescriptor}}, which stores location of 
> a {{BLOCKING_PERSISTENT}} {{ResultPartition}}, currently we only support 
> {{ResultPartition}} in {{TaskManager}}.
>  - Add a new method {{getResultPartitionDescriptors()}} in 
> {{AccessExecutionGraph}}
>  - Add a new filed in {{JobExecutionResult}}, {{JobResult}}, 
> {{ArchivedExecutionGraph}} and {{ExecutionEnvironment}}, which keeps a 
> mapping from {{IntermediateDataSetID}} to its {{ResultPartition}} locations
>  - When a job finishes, the metadata will flow in this path: 
> {{ExecutionGraph}} -> {{ArchivedExecutionGraph}} -> {{JobExecutionResult}} -> 
> {{JobResult}} -> {{ExecutionEnvironment}}



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

Reply via email to