[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17078030#comment-17078030
]
Piotr Nowojski commented on FLINK-13497:
The bug was finally fixed by combination of FLINK-16945
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16900790#comment-16900790
]
Biao Liu commented on FLINK-13497:
--
To [~pnowojski], yes, my pleasure :)
WRT the performance concern,
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16900776#comment-16900776
]
Piotr Nowojski commented on FLINK-13497:
[~SleePy] we are also not aware of any performance
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16900775#comment-16900775
]
Biao Liu commented on FLINK-13497:
--
To [~till.rohrmann], thanks for explanation of history.
I just
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16900764#comment-16900764
]
Till Rohrmann commented on FLINK-13497:
---
The reason why we don't run the {{CheckpointCoordinator}}
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16900758#comment-16900758
]
Biao Liu commented on FLINK-13497:
--
To [~pnowojski], I can't agree more.
I have another concern, the
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16900732#comment-16900732
]
Piotr Nowojski commented on FLINK-13497:
As far as we (me, [~carp84] and [~StephanEwen])
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16899896#comment-16899896
]
Till Rohrmann commented on FLINK-13497:
---
[~pnowojski] I think this issue would benefit from your
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16898822#comment-16898822
]
vinoyang commented on FLINK-13497:
--
[~SleePy] I discussed with [~yunta]. Yes, your analysis is correct.
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16898775#comment-16898775
]
Biao Liu commented on FLINK-13497:
--
The thread model of {{CheckpointCoordinator}} seems to be a bit
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16897896#comment-16897896
]
vinoyang commented on FLINK-13497:
--
[~SleePy] Your comment is reasonable, I am discussing with [~yunta]
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16897891#comment-16897891
]
Biao Liu commented on FLINK-13497:
--
There is another problem. The {{failGlobal}} might fail the job in
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16896988#comment-16896988
]
vinoyang commented on FLINK-13497:
--
When I thought deeply, even if we did not introduce FLINK-12364,
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16896773#comment-16896773
]
vinoyang commented on FLINK-13497:
--
[~yunta] I have no objection to stopping the checkpoint scheduler.
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16896767#comment-16896767
]
Yun Tang commented on FLINK-13497:
--
[~yanghua], there still exists a gap between fail all pending
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16896752#comment-16896752
]
vinoyang commented on FLINK-13497:
--
Currently, the {{CheckpointFailureManager}} choose a simple
[
https://issues.apache.org/jira/browse/FLINK-13497?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16896390#comment-16896390
]
Yun Tang commented on FLINK-13497:
--
One quick fix to this problem is to let
17 matches
Mail list logo