Help Center> Cloud Data Migration> Best Practices> Consulting Service Data Migration in Financial Industry> Using CDM to Upload Data to OBS> Uploading Inventory Data
View PDF

Uploading Inventory Data

  1. Use Direct Connect to establish a Direct Connect connection between the local data center and HUAWEI CLOUD Virtual Private Cloud (VPC).
  2. Create an OBS bucket and record the access domain name, port number, access key ID (AK), and secret access key (SK) of the OBS bucket.
  3. Log in to the CDM management console and click Buy CDM in the upper right corner to create a CDM cluster. See Figure 1.
    • Instance Type: Select cdm.medium, which applies to most migration scenarios.
    • VPC: VPC of the CDM cluster. Select the VPC that connects to the local data center through Direct Connect.
    • (Optional) Subnet and Security Group: You can configure either of them.
    • Auto Shutdown and Scheduled Startup/Shutdown: Retain the default values.
    Figure 1 Creating a cluster
  4. After the cluster is created, choose Job Management > Link Management > Create Link. The page for selecting a link type is displayed. See Figure 2.
    Figure 2 Selecting a connector
  5. To connect to the local Apache HDFS of company H, select Apache HDFS, and click Next. Set the HDFS link parameters as required. See Figure 3.
    Figure 3 Creating an HDFS link
    NOTE:
    • Name: Enter a custom link name, for example, hdfs_link.
    • URI: Enter the NameNode URI of HDFS of company H.
    • Authentication Method: Select Kerberos if Hadoop is in security mode to obtain the principal and keytab files from the client for authentication.
    • Principal and Keytab File: Obtain the principal account and keytab file from the Hadoop administrator.
  6. Click Save. CDM automatically checks whether the link is available.
    • If the link is available, a message is displayed, indicating that the link is successfully saved, and the link management page is displayed.
    • If the link is unavailable, check whether the link parameters are correctly configured or whether the firewall of company H allows the elastic IP address (EIP) of the CDM cluster to access the data source.
  7. Click Create Link to create an OBS link. On the page that is displayed, select HUAWEI CLOUD OBS, and click Next. Set the OBS link parameters as required. See Figure 4.
    Figure 4 Creating an OBS link
    NOTE:
    • Name: Enter a custom link name, for example, obslink.
    • OBS Server: Enter the domain name or IP address of OBS, for example, obs.myhuaweicloud.com.
    • Port: Enter the port number of OBS, for example, 443.
    • AK and SK: Enter the AK and SK used for accessing the OBS database. To obtain the AK and SK, log in to the management console, click the username in the upper right corner, and select My Credential from the drop-down list. On the page that is displayed, click the Access Keys tab.
  8. Click Save. The Link Management page is displayed.
  9. Choose Table/File Migration > Create Job to create a job for migrating trade statistics of company H to OBS. See Figure 5.
    Figure 5 Creating a job
    NOTE:
    • Job Name: Enter a user-defined job name.
    • Source Link Configuration:
      • Source Link Name: Select the HDFS link hdfs_link created in 5.
      • Source Directory/File: Set this parameter to the local storage path of company H's trade statistics. The value can be either a directory or a file. Set this parameter to a directory. CDM migrates all files in the directory to OBS.
      • File Format: Select Binary. The file format refers to the format used by CDM to transmit data. The formats of the original files are not changed. Binary is recommended for migration between files because the transmission efficiency and performance are optimal.
    • Destination Link Configuration:
      • Destination Link Name: Select the OBS link obslink created in 7.
      • Bucket Name and Write Directory: Enter the path for storing trade statistics in OBS. CDM writes the files to this path.
      • File Format: Select Binary. Similar to the source link, the formats of the original files are not changed.
      • Duplicate File Processing Method: Select Skip. CDM determines that a file is a duplicate file only when the file name and file size are the same on the source and destination ends. In this case, CDM skips the file and does not migrate the file to OBS.
  10. Click Next to go to the tab page for configuring the task parameters. For the migration of inventory data, retain the default values of the parameters.
  11. Click Save and Run. The Job Management page is displayed. You can view the job execution progress and result.
  12. After the job is successfully executed, click Historical Record to view the number of written rows, number of read rows, number of written bytes, number of written files, and execution logs.