If the current configurations for your cluster do not meet your requirements, you can change the configurations. Resources can be scaled up, down, in, or out to change your cluster configurations.

Prerequisites

  • The cluster does not have pending tasks for configuration changes.
  • Your Alibaba Cloud account does not have overdue renewal orders.

Background information

To change the configurations of an ApsaraDB for ClickHouse cluster, perform the following operations:
  • Scale-up or scale-down: Scale up or down node specifications, scale up node storage space, or scale up the computing capability of the cluster. For more information, see Scale-up or scale-down.
    Note
    • The configuration change process takes about 10 to 15 minutes to complete, during which the instance is restarted once and disconnected once for a few seconds. We recommend that you change configurations at off-peak hours.
    • When the performance metric of the cluster does not meet your business requirements, prioritize scale-up configurations.
  • Scale-out or scale-in: Increase or decrease the number of nodes. Scale out the computing capability of the cluster. For more information, see Scale-out or scale-in.

After you change the cluster configurations, the billing rules for the cluster may also change. For more information, see Billing details about specification changes.

Scale-up or scale-down

  1. Log on to the ApsaraDB for ClickHouse console.
  2. In the top navigation bar, select the region where the ApsaraDB for ClickHouse cluster for which you want to scale up or down resources is deployed.
  3. Click More in the Actions column that corresponds to the cluster and select Scale Up or Scale Down.
    Scale-up or scale-down
  4. Set Specifications for your node.
    Notice Specification and Storage Space configurations cannot be changed at the same time. However, you can change one parameter and then change the other after the first configuration takes effect.
  5. Click Buy Now and complete the payment. It takes about 5 to 10 minutes for the configuration changes to take effect.

Scale-out or scale-in

  1. Log on to the ApsaraDB for ClickHouse console.
  2. In the top navigation bar, select the region where the ApsaraDB for ClickHouse cluster for which you want to scale in or out resources is deployed.
  3. Click Scale Out or choose More > Scale In in the Actions column that corresponds to the cluster.
  4. If you click Scale Out, select the scale-out method and click Next. If you choose More > Scale In, this step is skipped, and the detection message appears. expand_mode
    • Migration Expansion

      If you select Migration Expansion, a new cluster is created and new nodes are added to it. Data of the original cluster is replicated to the new cluster. The data is written to shards of the new cluster based on the sharding key and the sharding method used by the original cluster. This ensures the load balancing among the shards after the cluster scale-out.

      Note The new cluster uses the same ID as the original cluster.
    • Simple Expansion

      If you select Simple Expansion, new nodes are added to the original cluster. Distributed and local tables are created on the new nodes. After the cluster scale-out, new data is written based on original distribution policies.

  5. In the scale-out or scale-in detection message, view the detection status. If you select Simple Expansion, detection is skipped, and the Upgrade/Downgrade page appears.
    • If the scale-out detection is complete, click Next.
      Notice
      • During the scaling action, the cluster enters the read-only state.
      • When the system migrates engine tables that are from the MergeTree family during the cluster scale-out, existing data is written to a new cluster and automatically redistributed.
      • When the system migrates engine tables that are from a non-MergeTree family such as external tables and log tables during the cluster scale-out, only table schemas are migrated. Data is not migrated. When the system migrates materialized views during the cluster scale-out, only table schemas are migrated. Data is not migrated.
      • After the cluster scale-out is complete, the IP address of the internal node changes. To write and query data by using an IP address of the node, you must obtain the node IP address again. For more information, see Obtain the CIDR block of the VPC where a cluster is deployed.
      Scale-out detection is complete
    • If an error occurred when the system performs detection for the scale-out, modify information and click Retry Detection. After the scale-out detection is complete, click Next.
      Causes of detection errors:
      • The unique distributed table is missing: No distributed tables are created for the local table. You must create a distributed table for the local table.
      • More than one distributed table exists for the local table: Delete redundant distributed tables until only one distributed table exists for the local table.
  6. On the Upgrade/Downgrade page, specify Nodes based on your requirements.
  7. Click Buy Now and complete the payment. The time that is required to complete the scale-out depends on the data volume.