Dataphin:Manage collection instances

Permission description

Super administrators, system administrators, and custom global roles with the Collection Task And Collection Instance-Management permission can manage collection instances.

Procedure

1. In the top menu bar of the Dataphin homepage, select Administration > Metadata.

2. In the left-side navigation pane, select Metadata Collection > Collection Instance to go to the Collection Instance page.

   The Collection Instance page retains records of instances that have run in the last 7 days, including the instance name, execution status, data source and data source encoding, data source type, runtime, execution method, executor, and task owner information.

   • Data Source: You can click the name of a data source to navigate to Management Hub > Data Source Management to view the information of the data source.

   • Execution Status: View the execution status of the corresponding instance in the collection instance list. The execution status can be Success, Failed, Running, Waiting, or Not Running/Paused, and applies only to a single operation. The operations that can be performed under each execution status are shown in the following table.

     Execution Status	Operations
     Success	Collection Change Overview, View Run Log, View Collection Task.
     Failed	Rerun, View Run Log, View Collection Task.
     Running	Stop, View Run Log, View Collection Task.
     Waiting	Stop, View Collection Task.
     Not Running/Paused	View Collection Task.

3. (Optional) You can search for a destination instance by instance name, or use quick filters to view task instances that you own, instances that you executed, or failed instances. You can also filter destination instances by execution status, data source type, data source environment, data source, task owner, executor, or execution method.

4. You can perform the following operations in the Actions column of a destination instance.

   Operation	Description
   Collection Change Overview	Click the icon in the Actions column. This shows the changes in the metadata checklist compared to the previous successful collection, including data source, collection scope, collection policy, object statistics, new object statistics, updated object statistics, deleted object statistics, table details, view details, and field details. Object Statistics: Shows the object statistics collected from the specified data source after this collection is successful. Change Overview List: Based on the collection policy, shows the additions, updates, and deletions of metadata compared to the previous successful collection.
   Rerun	Rerun a single instance: Click the icon in the Actions column to rerun a failed instance. Batch rerun: Select multiple instances or click Select All at the bottom, then click the icon at the bottom to rerun failed instances. Note Rerunning may take a long time. Please be patient.
   Stop	Only instances in the Running or Waiting state support the stop operation. Stop a single instance: Click the icon in the Actions column. Only instances in the Running or Waiting state support the stop operation. Batch stop: Select multiple instances or click Select All at the bottom, then click the icon at the bottom. Only instances in the Running or Waiting state support the stop operation. Note After being stopped, the execution status of the instance will be set to Failed. This is applicable to scenarios where the collection source cannot be connected, the collection task configuration is abnormal, or system exceptions cause the collection task to need early termination or cannot run normally. You can manually stop the task and then troubleshoot the exception through the run log.
   View Run Log	Click the icon in the Actions column to view the run log of the most recent instance. Each collection task corresponds to a workflow. You can search by node ID or node name, filter by run status, or use the quick filter to view Failed logs. You can also switch between different task types in the node list on the left to view their run logs: For Dlink type nodes, you can view log information, exception information, performance diagnostics, and run code. For other task types, you can only view log information. For failed nodes, you can click the Rerun button to rerun the instance. When the collection frequency of a collection task is set to Monthly, the system will generate a dry-run instance by default, mark it with a dry-run identifier, and set it directly to successful. For failed nodes, the system will rerun them according to the retry count configured in the collection task and mark them with an automatic retry identifier.
   View Collection Task	Click the instance name or the icon in the Actions column to navigate to the collection task list page. The system will filter out the collection task corresponding to the instance for you.