If the number of connections to an instance exceeds the upper limit or if an access exception occurs, you can restart the Tair（Redis OSS-compatible） instance.
Impacts and precautions
Data loss may occur during an instance restart.
During the restart process, the instance may experience transient connections. Before you restart the instance, assess the potential impact on your business and make sure that necessary preparations are in place. In addition, make sure that your applications can reconnect to the instance.
Procedure
Log on to the ApsaraDB for Redis console and go to the Instances page. In the top navigation bar, select the region in which the instance that you want to manage resides.
Find the instance that you want to restart and click Restart in the Actions column.
In the dialog box that appears, configure Restart mode.
Restart Now: immediately restarts the instance.
Restart Within Maintenance Window: restarts the instance within the specified maintenance window. For more information, see Configure a maintenance window.
Read and select the risks of restarting an instance. Then, click OK.
FAQ
After I restart an instance, the instance remains in the Restarting state. Why?
After you restart an ApsaraDB for Redis Community Edition or ApsaraDB for Redis Enhanced Edition (Tair) instance, the instance state immediately changes to Restarting.
If you select Restart Within Maintenance Window when you restart the instance, the instance performs a master-replica switchover or shard failover within the maintenance window. Before the maintenance window begins, the instance can provide services normally.
How long does it take to restart an instance?
It usually takes no more than 10 minutes to restart an instance.
Does an instance restart lead to data loss?
The primary objective of restarting an instance is to quickly restore service availability.
Restarting a high-availability master-replica instance generally does not result in data loss. However, if the synchronization between the master and replica nodes is delayed due to reasons such as heavy read/write operations before the instance is restarted, the data that has not been synchronized to the replica node may be lost.
If you restart a standalone instance, data loss may occur. Exercise caution when you restart the instance. If your business requires high data reliability, we recommend that you upgrade the standalone instance to a high-availability master-replica instance.