Advantages of the RDS

During the use of the RDS, you need not to worry about the database’s basic OPS work (like high availability and data reliability) and database clustering work (like read-write separation and multi-data center). You just keep a close eye on your services and check if the RDS instance’s performance is adequate. If not, you can optimize your services by upgrading instance, adding read-only replica, or analyzing slow query.

Instructions on Risks from the RDS Configuration Change

The configuration change of the RDS instance involves the current instance’s hot backup node and a pair of master and hot backup nodes that are newly generated. Also, the change does not influence the current master nodes. Despite this, we still advise you to perform the operations at the low peak of traffic. Since it takes some time to change the configuration of the RDS to ensure the data integrity and consistency, we import the incremental data generated in this period into a pair of newly-generated master-backup nodes. However, massive service traffic might lead to overmuch incremental data and bring some uncertain risks. After the configuration change, the instance is subject to restarting and entry switching, so connection failure at the second level might take place. Recommend you to make preparations and configure the reconnection between service and server in advance.

Instructions on Risks from Data Recovery of the RDS

Presently, the RDS’s data recovery feature only refers to the complete replacement of existing instance data with snapshot data. Before executing the data recovery, recommend you to repeatedly confirm the recovery and back up important data to prevent the data recovery from bringing about data loss. If you need to roll b ack only partial data, recommend you to generate one new RDS instance via the “Clone Instance” feature. In the new RDS instance, you should make sure the data has no problems and then import the desired data into the production library.

Instructions on Risks from Disk Capacity of the RDS

Presently, the RDS instance’s space of use only calculates the user’s data use space, and it cannot calculate the log and temporary table. Herein, temporary tables, which are generated during instance running, are stored in complimentary disk space. However, using a large temporary table during running (the required temporary space is higher than 200% of the overall instance’s residual space) can result in full occupation of the instance disk and service unavailability. In this case, you need to close the link using the large temporary table and release the disk space. Or, you can wait till the automated recovery module switches your database entry to the hot backup node.

Instructions on Fault Switch of RDS

The RDS master instance adopts the “main node +hot backup node” high-availability mode. If the main node is disabled, the automated recovery module switches your database entry to the hot backup node within one minute. However, you should know this in advance and configure the reconnection between the service and server. Thus, it avoids service unavailability resulting from the change of entry.