To scale out or scale up an instance, you can add nodes to the instance or increase the storage capacity of the data disks that are used by the instance. In the ApsaraDB for HBase console, the Add Node and Expand Disk buttons in the Actions column are available for each instance, as shown in the following figure.
You can add nodes to an instance. The new nodes have the same configuration as the existing nodes. For example, the number of CPU cores, memory size, disk type, and disk capacity are the same as those of the existing nodes. You can add up to five nodes at a time. An ApsaraDB for HBase instance can contain up to 100 nodes. If your business requires more than 100 nodes, submit a ticket.
When you add nodes to an instance, new disks are also added to the instance. This increases the storage capacity of the instance.
Impacts on the instance: In normal conditions, instance scale-out operations cause controllable impacts. However, we recommend that you scale out your ApsaraDB for HBase instances during off-peak periods. After new nodes are added to an instance, ApsaraDB for HBase runs the server load balancing service. During this process, some data partitions are transferred to the new nodes. When a partition is being transferred, your workloads on that partition are interrupted for a short period of time. In normal conditions, the interruption is within 200 milliseconds. If the retry feature is enabled on your client, the client retries the interrupted operations. By default, the retry feature is enabled. However, if you configure a short timeout period for the client, request errors may occur.
Increase the capacity of disks
You can increase the capacity of the data disks that are used by an instance. You can increase the capacity of each disk by at least 40 GB. Each disk can have a capacity of 2,000 GB.
Impacts on the instance: After you increase the capacity of the disks, the system performs a rolling restart on the RegionServers and the DataNodes. This has minor impacts on the instance.