[ 
https://issues.apache.org/jira/browse/FLINK-16536?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Zhijiang updated FLINK-16536:
-----------------------------
    Description: 
During recovery process for unaligned checkpoint, the input channel state 
should also be recovered besides with existing operator states.

The InputGate would request buffer from local pool and then interact with 
ChannelStateReader to fill in the state data.  The filled buffer would be 
inserted into respective InputChannel queue for processing in normal way.

It should guarantee that the new data from upstream side should not overtake 
the input state data to avoid mis-order issue.

Refer to more details by 
[https://docs.google.com/document/d/16_MOQymzxrKvUHXh6QFr2AAXIKt_2vPUf8vzKy4H_tU/edit]

  was:
During recovery process for unaligned checkpoint, the input channel state 
should also be recovered besides with existing operator states.

The InputGate would request buffer from local pool and then interact with 
ChannelStateReader to fill in the state data.  The filled buffer would be 
inserted into respective InputChannel queue for processing in normal way.

It should guarantee that the new data from upstream side should not overtake 
the input state data to avoid mis-order issue.

Refer to more details by 
[https://docs.google.com/document/d/16_MOQymzxrKvUHXh6QFr2AAXIKt_2vPUf8vzKy4H_tU/edit]]


> Implement InputChannel state recovery for unaligned checkpoint
> --------------------------------------------------------------
>
>                 Key: FLINK-16536
>                 URL: https://issues.apache.org/jira/browse/FLINK-16536
>             Project: Flink
>          Issue Type: Sub-task
>          Components: Runtime / Checkpointing, Runtime / Network
>            Reporter: Zhijiang
>            Priority: Major
>             Fix For: 1.11.0
>
>
> During recovery process for unaligned checkpoint, the input channel state 
> should also be recovered besides with existing operator states.
> The InputGate would request buffer from local pool and then interact with 
> ChannelStateReader to fill in the state data.  The filled buffer would be 
> inserted into respective InputChannel queue for processing in normal way.
> It should guarantee that the new data from upstream side should not overtake 
> the input state data to avoid mis-order issue.
> Refer to more details by 
> [https://docs.google.com/document/d/16_MOQymzxrKvUHXh6QFr2AAXIKt_2vPUf8vzKy4H_tU/edit]



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to