[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16715414#comment-16715414
]
stack commented on HBASE-21559:
---
This cleaned up the failures nicely. Thanks [~openinx].
> The
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16712691#comment-16712691
]
Hudson commented on HBASE-21559:
Results for branch master
[build #650 on
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16712675#comment-16712675
]
Hudson commented on HBASE-21559:
Results for branch branch-2
[build #1544 on
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16712338#comment-16712338
]
stack commented on HBASE-21559:
---
Nice fix [~openinx]. Running it locally, I can't make the hang anymore.
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16712323#comment-16712323
]
Hudson commented on HBASE-21559:
Results for branch branch-2.1
[build #664 on
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16712311#comment-16712311
]
Hudson commented on HBASE-21559:
Results for branch branch-2.0
[build #1143 on
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16712174#comment-16712174
]
Duo Zhang commented on HBASE-21559:
---
Pushed to branch-2.0+. Let's see how it works.
> The
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711921#comment-16711921
]
Hadoop QA commented on HBASE-21559:
---
| (/) *{color:green}+1 overall{color}* |
\\
\\
|| Vote ||
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711604#comment-16711604
]
Hadoop QA commented on HBASE-21559:
---
| (/) *{color:green}+1 overall{color}* |
\\
\\
|| Vote ||
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711538#comment-16711538
]
Zheng Hu commented on HBASE-21559:
--
bq. But at t8, the TakeSnapshotHandler is already in the map right?
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711506#comment-16711506
]
Duo Zhang commented on HBASE-21559:
---
But at t8, the TakeSnapshotHandler is already in the map right?
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711503#comment-16711503
]
Zheng Hu commented on HBASE-21559:
--
bq. But they will not hold the map lock when modifying right?
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711436#comment-16711436
]
Zheng Hu commented on HBASE-21559:
--
bq. Do we really need to test the isFinished under the map lock?
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711430#comment-16711430
]
Duo Zhang commented on HBASE-21559:
---
Do we really need to test the isFinished under the map lock? Just
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711439#comment-16711439
]
Duo Zhang commented on HBASE-21559:
---
But they will not hold the map lock when modifying right?
> The
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711429#comment-16711429
]
Zheng Hu commented on HBASE-21559:
--
BTW, I think we can move the snapshot feature from procedure.v1
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711425#comment-16711425
]
Zheng Hu commented on HBASE-21559:
--
Run the UT in my localhost 5 times, Seems OK now.
> The
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711397#comment-16711397
]
Zheng Hu commented on HBASE-21559:
--
Currently, the snapshotManager grab the object lock in many
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711212#comment-16711212
]
Duo Zhang commented on HBASE-21559:
---
Oh we have some logic in SnapshotProcedure and
[
https://issues.apache.org/jira/browse/HBASE-21559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16711214#comment-16711214
]
Zheng Hu commented on HBASE-21559:
--
Yeah, It's a dead lock, The SplitTableRegionProcedure grabed the
20 matches
Mail list logo
