Mailing-List: contact hdfs-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: hdfs-issues@hadoop.apache.org
Date: Wed, 25 Nov 2015 07:18:11 +0000 (UTC)
From: =?utf-8?Q?=E5=88=98=E5=96=86_=28JIRA=29?= <jira@apache.org>
To: hdfs-issues@hadoop.apache.org
Message-ID: <JIRA.12746626.1412748975000.188246.1448435891019@Atlassian.JIRA>
In-Reply-To: <JIRA.12746626.1412748975000@Atlassian.JIRA>
References: <JIRA.12746626.1412748975000@Atlassian.JIRA>
 <JIRA.12746626.1412748975497@arcas>
Subject: [jira] [Commented] (HDFS-7208) NN doesn't schedule replication when
 a DN storage fails
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable


    [ https://issues.apache.org/jira/browse/HDFS-7208?page=3Dcom.atlassian.=
jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D15026=
339#comment-15026339 ]=20

=E5=88=98=E5=96=86 commented on HDFS-7208:
--------------------------

We meet the same problem, but we have a very simple path that works.  We ca=
n treat it as the datanode deleted the related blocks, so we only need one =
line to fix it.


diff --git a/hadoop/adh/src/hadoop-hdfs-project/hadoop-hdfs/src/main/java/o=
rg/apache/hadoop/hdfs/server/datanode/fsdataset/impl/FsDatasetImpl.java b/h=
adoop/adh/src/hadoop-hdfs-proje
index 3320c65..7a10072 100644
--- a/hadoop/adh/src/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apac=
he/hadoop/hdfs/server/datanode/fsdataset/impl/FsDatasetImpl.java
+++ b/hadoop/adh/src/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apac=
he/hadoop/hdfs/server/datanode/fsdataset/impl/FsDatasetImpl.java
@@ -1332,6 +1332,8 @@ public void checkDataDir() throws DiskErrorException =
{
                   + " on failed volume " + fv.getCurrentDir().getAbsoluteP=
ath());
               ib.remove();
               removedBlocks++;
+              datanode.notifyNamenodeDeletedBlock(new ExtendedBlock(bpid, =
b.getBlockId()), b.getStorageUuid());
             }
           }
         }

> NN doesn't schedule replication when a DN storage fails
> -------------------------------------------------------
>
>                 Key: HDFS-7208
>                 URL: https://issues.apache.org/jira/browse/HDFS-7208
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: namenode
>            Reporter: Ming Ma
>            Assignee: Ming Ma
>             Fix For: 2.6.0
>
>         Attachments: HDFS-7208-2.patch, HDFS-7208-3.patch, HDFS-7208.patc=
h
>
>
> We found the following problem. When a storage device on a DN fails, NN c=
ontinues to believe replicas of those blocks on that storage are valid and =
doesn't schedule replication.
> A DN has 12 storage disks. So there is one blockReport for each storage. =
When a disk fails, # of blockReport from that DN is reduced from 12 to 11. =
Given dfs.datanode.failed.volumes.tolerated is configured to be > 0, NN sti=
ll considers that DN healthy.
> 1. A disk failed. All blocks of that disk are removed from DN dataset.
> =20
> {noformat}
> 2014-10-04 02:11:12,626 WARN org.apache.hadoop.hdfs.server.datanode.fsdat=
aset.impl.FsDatasetImpl: Removing replica BP-1748500278-xx.xx.xx.xxx-137780=
3467793:1121568886 on failed volume /data/disk6/dfs/current
> {noformat}
> 2. NN receives DatanodeProtocol.DISK_ERROR. But that isn't enough to have=
 NN remove the DN and the replicas from the BlocksMap. In addition, blockRe=
port doesn't provide the diff given that is done per storage.
> {noformat}
> 2014-10-04 02:11:12,681 WARN org.apache.hadoop.hdfs.server.namenode.NameN=
ode: Disk error on DatanodeRegistration(xx.xx.xx.xxx, datanodeUuid=3Df3b8a3=
0b-e715-40d6-8348-3c766f9ba9ab, infoPort=3D50075, ipcPort=3D50020, storageI=
nfo=3Dlv=3D-55;cid=3DCID-e3c38355-fde5-4e3a-b7ce-edacebdfa7a1;nsid=3D420527=
250;c=3D1410283484939): DataNode failed volumes:/data/disk6/dfs/current
> {noformat}
> 3. Run fsck on the file and confirm the NN's BlocksMap still has that rep=
lica.


--
This message was sent by Atlassian JIRA
(v6.3.4#6332)