Trafodion Troubleshooting-Failed to replace a bad datanode on the existing pipeline

最新推荐文章于 2023-05-19 11:43:55 发布

原创最新推荐文章于 2023-05-19 11:43:55 发布 · 552 阅读

1 ·

CC 4.0 BY-SA版权

QianBase/EsgynDB/Trafodion 同时被 2 个专栏收录

263 篇文章

订阅专栏

QianBase/EsgynDB诊断

65 篇文章

订阅专栏

在单节点上部署EsgynDB时遇到初始化错误，错误信息显示为LOB_DATA_WRITE_ERROR_RETRY。通过调整HDFS配置参数并设置dfs.replication为1，成功解决了问题。

现象

安装EsgynDB在单个节点，在数据库初始化过程中报错，错误内容如下，

Create Library Manager: Started

*** ERROR[8458] Unable to access ExpLOBInterfaceInsert interface after retry. Call to ExpLOBInterfaceInsert returned error LOB_DATA_WRITE_ERROR_RETRY(517). Error detail: 0. Cause: java.io.IOException: IOException flush: java.io.IOException: Failed to replace a bad datanode on the existing pipeline due to no more good datanodes being available to try. (Nodes: current=[DatanodeInfoWithStorage[192.168.20.117:50010,DS-7705fe88-0f38-4cf5-afaa-96970779ccac,DISK]], original=[DatanodeInfoWithStorage[192.168.20.117:50010,DS-7705fe88-0f38-4cf5-afaa-96970779ccac,DISK]]). The current failed datanode replacement policy is DEFAULT, and a client may configure this via 'dfs.client.block.write.replace-datanode-on-failure.policy' in its configuration.
org.apache.hadoop.hdfs.DFSOutputStream.flushOrSync(DFSOutputStream.java:2532)
org.apache.hadoop.hdfs.DFSOutputStream.hflush(DFSOutputStream.java:2378)
org.apache.hadoop.fs.FSDataOutputStream.hflush(FSDataOutputStream.java:130)
org.trafodion.sql.HDFSClient.hdfsWriteImmediate(HDFSClient.java:567).

解决

网上搜索方案如下，
修改hdfs-site.xml文件，添加或者修改如下两项：

<property>
<name>dfs.client.block.write.replace-datanode-on-failure.enable</name>
<value>true</value>
</property>
<property>
<name>dfs.client.block.write.replace-datanode-on-failure.policy</name>
<value>NEVER</value>
</property>

添加以上参数并重启后发现错误依然存在，检查HDFS参数dfs.replication值为3，即设置为3副本。但因为只有单个节点，设置dfs.replication=1并重启HDFS，问题解决！