As your business develops, you may have higher requirements for the configuration of your Elasticsearch cluster. If the current configuration of your Elasticsearch cluster cannot meet your business needs, you can upgrade its configuration. This topic describes how to upgrade the configuration of an Elasticsearch cluster and the related precautions.
- You may not be able to change some of the preceding items due to limits. For more information, see Procedure.
- If the zone where your Elasticsearch cluster resides has insufficient resources for a configuration upgrade, you can migrate the nodes in the zone to another before the upgrade. For more information, see Migrate nodes in a zone.
- Changes in specifications
- For each upgrade, you can change the configuration for only one type of node. The node types are data nodes, warm nodes, client nodes, dedicated master nodes, and Kibana nodes.
- You cannot change the disk types of nodes when you upgrade the configuration of your cluster. You can only change the storage space of a single node.
- You can only downgrade your cluster by removing data nodes. The number of data nodes that you can remove is limited. For more information, see Scale in an Elasticsearch cluster. Currently, you cannot reduce disk space or downgrade nodes.
- Impact on services
- If your cluster is abnormal (indicated by the color yellow or red), you must select Forced Update to make changes take effect. This may affect services.
- In most cases, the system restarts your cluster after a configuration upgrade to make changes take effect. However, if your cluster has dedicated master nodes and you only change the number of nodes, the system does not restart the cluster.
- Changes in the cluster version
Currently, you can only upgrade your Elasticsearch clusters from V6.3.2 to V6.7.0. For more information, see Upgrade the version of a cluster.
- Changes in billing
- You can upgrade the Kibana node. However, this upgrade incurs fees.
- After you submit a configuration upgrade order, your cluster is billed based on the
Note During a configuration upgrade, you can check the price of your order on the Update page in real time.
- Log on to the Alibaba Cloud Elasticsearch console.
- In the top navigation bar, select the region where your Alibaba Cloud Elasticsearch cluster resides.
- Find the target cluster and click its ID.
- In the lower-right corner of the Basic Information page, click Upgrade.
- On the Update page, change the configuration of the cluster based on the following instructions.
Note The Current Config section on the Update page shows the current configuration of the cluster. You can use this as a reference when you perform the upgrade.
Follow the instructions on the Update page to upgrade the configuration of your cluster based on your business needs. For more information about the parameters on the Update page, see Parameters. The following table describes only some of the parameters.
Parameter Description Dedicated Master Node You can upgrade purchased dedicated master nodes. By default, the number of dedicated master nodes is 3, and each dedicated master node offers 2 vCPUs, 8 GiB of memory, and a cloud disk with 20 GiB. After the upgrade, the cluster is billed based on the new configuration.Note If you have purchased dedicated master nodes with the specifications of one vCPU and 2 GiB of memory, you can upgrade these nodes on the Update page. After the upgrade, the cluster is billed based on the new configuration. If your dedicated master nodes are free of charge, they incur fees after you upgrade them. Client Node You can upgrade purchased client nodes. By default, the number of client nodes is 2, and each client node offers 2 vCPUs, 8 GiB of memory, and a cloud disk with 20 GiB. After the upgrade, the cluster is billed based on the new configuration. Warm Node You can upgrade purchased warm nodes. By default, the number of warm nodes is 2, and each warm node offers 2 vCPUs, 8 GiB of memory, and a cloud disk with 500 GiB. After the upgrade, the cluster is billed based on the new configuration. Kibana Node You can upgrade the purchased Kibana node. By default, the Kibana node offers two vCPUs and 4 GiB of memory.Notice When you purchase a cluster, a free-of-charge Kibana node with one vCPU and 2 GiB of memory is provided. After you upgrade the Kibana node, the cluster is billed based on the new configuration. Forced Update If your cluster is abnormal (indicated by the color red or yellow), your services are severely affected, and you must upgrade the cluster configuration immediately, select Forced Update. The system will perform a forced update regardless of the cluster status. The update takes only a short period of time.Notice
- After a forced update, the system restarts your cluster.
- During the restart of a cluster, the services running on the cluster may be unstable.
- If you do not select Forced Update, the system uses the default mode to restart your cluster to make the changes take effect. For more information, see Manage clusters.
- If your cluster is abnormal (indicated by the color red or yellow), the system automatically selects Forced Update. In this case, the system does not use the default mode to restart the cluster.
The storage space of nodes is measured in GiB. A standard SSD provides a maximum of 2,048 GiB of storage space.
You can resize an existing ultra disk to a maximum of 2,048 GiB. When you purchase an ultra disk, you can set the storage space to a maximum of 5,120 GiB. If the volume of the data you want to store exceeds 2,048 GiB, you can set the space to 2560, 3072, 3584, 4096, 4608, or 5120.
- Read and agree to the terms of cluster configuration upgrades. Then, click Buy Now and complete the payment as prompted.
After you complete the payment, the system restarts the cluster to make the changes take effect.