Troubleshooting Hadoop service failure
If any of the Hadoop services failed, then try Stop and Start the Hadoop services from the Service Manager. Even after restarting, if problem occurred, then check any of the following solutions,
S No | Issue | Cause | Fix | ||||||||||||
1 | Name node failed to start |
“$InstalledPath \BigDataSDK\Metadata\data\dfs\namenode”
| Steps to format Name node via Service Manager:
Note: Formatting Name node will clear all the data in your cluster.
| ||||||||||||
2 | Data node failure | Cluster ID version of Name node and Data node mismatch
| Formatting Name node will be the solution for this case. Refer the steps mentioned in #1 to format the Name node.
Note: Reason for Cluster ID mismatch:
Steps to ensure the Cluster ID in Name node and in Data node:
Version file Path: $InstalledPath\BigDataSDK\Metadata\data\dfs\datanode\current\Version (check clusterID in the file) $InstalledPath\BigDataSDK\Metadata\data\dfs\ namenode \current\Version (check clusterID in the file)
| ||||||||||||
3 | Service Manager shows alert while starting Hadoop services.
“Syncfusion Big Data Agent is not running. Please restart it in Windows services to manage Hadoop service.”
| Syncfusion Big Data Agent service may be stopped or uninstalled.
Note: Big Data Studio application requires 2 dependent window services listed below for its proper functioning,
| To start “Syncfusion Big Data Agent” service in the system. It can be done in below mentioned two different ways,
From Task Manager:
(or)From Command Prompt:
“ServiceName start” E.g.: SyncfusionBigDataAgent.exe start If the window services are not properly running, restart the system once and try to start the service again from services window.
Then try to start the Hadoop services from the Service Manager.
| ||||||||||||
4 | BindException | If a port on a machine is already in use and a Hadoop services (Name node, Job tracker, Data node, Task tracker etc.) tries to create its service on that same port.
Port details of the Hadoop services:
Note: You can find the exception details under the below specified path with the service name as a log file name “$InstalledPath \BigDataSDK\Metadata\”.
|
| ||||||||||||
5 | Insufficient disk space | Insufficient disk space in the drive which has the Hadoop “data” directory.
Directory path as follows, “$InstalledPath \BigDataSDK\Metadata\data”
|
|