Help Center > > User Guide> FusionInsight Manager Operation Guide (Applicable to 3.x)> Cluster> Cluster Management> Client Management> Batch Upgrading Clients

Batch Upgrading Clients

Updated at: Mar 25, 2021 GMT+08:00

Scenario

The client package downloaded from FusionInsight Manager contains the client batch upgrade tool. If multiple clients need to be upgraded after a cluster upgrade or capacity expansion, you can use this tool to batch upgrade the clients. In addition, the tool provides a lightweight function for batch updating the /etc/hosts file on the nodes where the clients are located.

Procedure

Preparation Before the Client Upgrade

  1. Log in to FusionInsight Manager.
  2. Choose Cluster > Name of the desired cluster > More > Download Client to download the client package to the specified directory on the server.

    For details, see section Downloading the Client.

    Decompress the downloaded client package and find the batch_upgrade directory, for example, /tmp/FusionInsight-Client/FusionInsight_Cluster_1_Services_ClientConfig/batch_upgrade.

  1. Choose Cluster > Name of the desired cluster > Client Management. On the Client Management page that is displayed, and click Export All to export the information about the selected clients to the local PC.
  2. Decompress the exported client information and upload the client-info.cfg file to the batch_upgrade directory.
  3. Supplement the ciphertext password in the client-info.cfg file by referring to Reference.

Batch Upgrading Clients

  1. Run the sh client_batch_upgrade.sh -u -f /tmp/FusionInsight-Client/FusionInsight_Cluster_1_Services_Client.tar -g /tmp/FusionInsight-Client/FusionInsight_Cluster_1_Services_ClientConfig/batch_upgrade/client-info.cfg command to perform the upgrade.

    Because the password is configured, you are advised to delete the client-info.cfg file as soon as possible after the upgrade.

  2. After the upgrade is complete, verify the upgrade result by running the sh client_batch_upgrade.sh -c command.
  3. If the client is faulty after the upgrade, run the sh client_batch_upgrade.sh -s command to roll back the upgrade.

    • The client batch upgrade tool moves the original client to the backup directory, and then uses the client package specified by the -f parameter to install the client again. Therefore, if the original client contains customized contents, manually save the customized contents from the backup directory or move the customized contents to the client directory after the upgrade before you run the -c command. Client backup path: {Original client path}-backup
    • The -u parameter is the prerequisite for the -c and -s commands. You can run the -c command to submit or the -s command to perform a rollback only after the -u command is executed to perform an upgrade.
    • You can run the -u command multiple times to upgrade only the clients that fail to be upgraded.
    • The client batch upgrade tool also supports clients of early versions.
    • To upgrade a client installed by a non-root user, the operator must have the read/write permission for the directory where the client is located and its parent directory. Otherwise, the upgrade fails.
    • The client package of the -f parameter must be a full client. The client package of a single component or some components cannot be used as the input.

Reference

Before upgrading clients in batches, you need to manually configure the user password for remotely logging in to the client node.

Run the vi client-info.cfg command to add the user password.

For example:

clientIp,clientPath,user,password
10.10.10.100,/home/omm/client /home/omm/client2,omm,Bigdata_123

The fields in the configuration file are as follows:

  • clientIp: Indicates the IP address of the node where the client is located.
  • clientPath: Indicates the client installation path. Multiple paths are separated by spaces. Note that the path cannot end with a slash (/).
  • user: Indicates the username of the node.
  • password: Indicates the user password of the node.
    • Enter a plaintext password.
    • If the execution fails, you can check node.log in work_space/log_XXX under the execution directory.

Did you find this page helpful?

Submit successfully!

Thank you for your feedback. Your feedback helps make our documentation better.

Failed to submit the feedback. Please try again later.

Which of the following issues have you encountered?







Please complete at least one feedback item.

Content most length 200 character

Content is empty.

OK Cancel