[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15440121#comment-15440121
]
Anu Engineer commented on HDFS-10598:
-
The issue is now tracked in HDFS-10808
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Anu Engineer
>Priority: Critical
> Fix For: 3.0.0-alpha2
>
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15439932#comment-15439932
]
Anu Engineer commented on HDFS-10598:
-
[~andrew.wang] Sure, I will open another JIRA and resolve this.
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Anu Engineer
>Priority: Critical
> Fix For: 3.0.0-alpha2
>
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15439859#comment-15439859
]
Andrew Wang commented on HDFS-10598:
[~anu] do you mind doing this work in another JIRA? Would prefer to avoid
addendum patches.
Or, we can revert and re-fix, but if we're pursuing a different approach, a new
JIRA feels better.
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Anu Engineer
>Priority: Critical
> Fix For: 3.0.0-alpha2
>
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15439842#comment-15439842
]
Anu Engineer commented on HDFS-10598:
-
[~eddyxu] First of all my apologies for missing this JIRA. I was on vacation
and this was queued up for later processing. [~arpitagarwal] Thanks for
flagging this and bringing it back to my attention.
[~jojochuang] and [~eddyxu] Thanks for working on this patch.
I think we have some subtle issues with this patch. First of all this patch
attempts to remove the use of SetExitFlag from copy blocks, but it is trying to
use the same in the finally block. The easier/correct fix is to make no
changes to copyBlocks, but just add a clearExistFlag in the run function. I
will make that change and post a patch soon.
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Fix For: 3.0.0-alpha2
>
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15438843#comment-15438843
]
Arpit Agarwal commented on HDFS-10598:
--
Thank you for the review and commit [~jojochuang]! I've been busy with other
commitments and never got around to reviewing this.
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Fix For: 3.0.0-alpha2
>
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15394758#comment-15394758
]
Hudson commented on HDFS-10598:
---
SUCCESS: Integrated in Hadoop-trunk-Commit #10158 (See
[https://builds.apache.org/job/Hadoop-trunk-Commit/10158/])
HDFS-10598. DiskBalancer does not execute multi-steps plan. Contributed
(weichiu: rev d84ab8a5786c9320e4708b2f0d54f978fee5ba95)
*
hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/DiskBalancer.java
*
hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/diskbalancer/TestDiskBalancer.java
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Fix For: 3.0.0-alpha2
>
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15394349#comment-15394349
]
Wei-Chiu Chuang commented on HDFS-10598:
+1. Given this is a critical fix, I will commit by end of day unless someone
wants to veto it. We could file a follow-up jira for the regression test if
necessary.
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15392760#comment-15392760
]
Wei-Chiu Chuang commented on HDFS-10598:
Hi [~eddyxu] thanks identifying the bug and submitting the patch. I think the
fix is straightforward and the unit test makes sense to me. I wonder if we need
more unit tests to cover more scenarios, because in addition to the normal
operation, the patch fixes the termination condition in these corner cases:
* {code}// Check for the max error count constraint.{code}
* {code}// we are not able to find any blocks to copy.{code}
* {code}// check if someone told us exit{code}
* {code}// Technically it is possible for us to find a smaller block and{code}
[~arpitagarwal], what's your take?
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15382630#comment-15382630
]
Arpit Agarwal commented on HDFS-10598:
--
Hi [~eddyxu], I will review in a couple of days as the fix needs some thought.
Last and this week are very busy for me.
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15382604#comment-15382604
]
Lei (Eddy) Xu commented on HDFS-10598:
--
Ping [~arpitagarwal] , would you mind to give a review on this patch? Thanks!
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15376006#comment-15376006
]
Lei (Eddy) Xu commented on HDFS-10598:
--
Great. Thanks Arpit .
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15375848#comment-15375848
]
Arpit Agarwal commented on HDFS-10598:
--
Disk Balancer is not in branch-2 so I've set updated the versions accordingly.
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Commented] (HDFS-10598) DiskBalancer does not execute multi-steps plan.
[
https://issues.apache.org/jira/browse/HDFS-10598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15375844#comment-15375844
]
Arpit Agarwal commented on HDFS-10598:
--
Hi [~eddyxu], thanks for reporting this problem and posting a patch. I believe
Anu is out on vacation for the next few weeks. I will review your fix.
> DiskBalancer does not execute multi-steps plan.
> ---
>
> Key: HDFS-10598
> URL: https://issues.apache.org/jira/browse/HDFS-10598
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: diskbalancer
>Affects Versions: 2.8.0, 3.0.0-beta1
>Reporter: Lei (Eddy) Xu
>Assignee: Lei (Eddy) Xu
>Priority: Critical
> Attachments: HDFS-10598.00.patch
>
>
> I set up a 3 DN node cluster, each one with 2 small disks. After creating
> some files to fill HDFS, I added two more small disks to one DN. And run the
> diskbalancer on this DataNode.
> The disk usage before running diskbalancer:
> {code}
> /dev/loop0 3.9G 2.1G 1.6G 58% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 17M 3.6G 1% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> However, after running diskbalancer (i.e., {{-query}} shows {{PLAN_DONE}})
> {code}
> /dev/loop0 3.9G 1.2G 2.5G 32% /mnt/data1
> /dev/loop1 3.9G 2.6G 1.1G 71% /mnt/data2
> /dev/loop2 3.9G 953M 2.7G 26% /mnt/data3
> /dev/loop3 3.9G 17M 3.6G 1% /mnt/data4
> {code}
> It is suspicious that in {{DiskBalancerMover#copyBlocks}}, every return does
> {{this.setExitFlag}} which prevents {{copyBlocks()}} be called multiple times
> from {{DiskBalancer#executePlan}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
