1.1.1. Product Consultation
What should I pay attention to when using RDS-PostgreSQL?
1. The operating system of the RDS-PostgreSQL instance is not visible to users. Only the user application can access the IP address and port corresponding to the database. Alternatively, you can bind an EIP and use it for access.
2. View the list of instances and make sure that it is consistent with the resource pool selected for the purchase instance.
3. After applying for an RDS-PostgreSQL instance, basic operation and maintenance of the database (such as resource monitoring, high availability, etc.) are not required. However, you need to pay attention to the following things:
• Check for performance issues related to the RDS-PostgreSQL instance. For instance, whether there are a great number of slow SQLs, and whether SQL statements need to be optimized, etc.
• Please check whether the RDS-PostgreSQL instance has sufficient CPU cores, memory, IOPS, storage space, and number of connections. If not, please optimize or update it.
What is RDS-PostgreSQL availability?
RDS-PostgreSQL instance availability formula: Instance availability = (1-fault time/total service time)×100%.
Will an RDS-PostgreSQL instance be affected by another user instance?
The RDS-PostgreSQL instances do not affect each other. The resources of instances are isolated and independent of each instance without mutual effect.
Does RDS-PostgreSQL support cross-AZ high availability?
Yes, RDS-PostgreSQL supports cross-AZ high availability. To prevent data loss and minimize cluster downtime in the event of service interruption, you can select one or two AZs in the same region when creating a cluster. The system will allocate nodes between the selected AZs.
Availability Zone (AZ) refers to the physical area in which the power and network are independent of each other within the same region. Intranet interoperability is available across AZs and different AZs are physically isolated. The RDS-PostgreSQL service supports the deployment of database primary/standby instances within the same AZ or across AZs.
If the primary instance and the standby instance are distributed in the same AZ, it is unable to ensure high availability when an AZ-level fault occurs. As such, it is recommended to deploy the primary/standby instance nodes separately.
Does RDS-PostgreSQL support changing a primary/standby instance to a single instance?
RDS-PostgreSQL only supports the expansion and upgrade of a single instance to a primary/standby instance. However, a primary/standby instance cannot be reduced to a single instance. To change a primary/standby instance to a single instance, it is recommended that you migrate the data of the primary/standby instance to a single instance through data migration.
Why application access is slower after using RDS-PostgreSQL?
Cloud data performance and access speed are related to multiple factors. We recommend that you perform the following checks:
• You can view the memory, CPU, IO, disk, QPS, and other performance data of an instance through the monitoring module of the Console of the RDS-PostgreSQL service.
• This issue may be related to the application. You can view the statement execution of the instance, slow SQL, and other information through the monitoring module of the Console of the RDS-PostgreSQL service.
• Or, this issue may be related to network conditions. You can compare the network differences between the application to the RDS-PostgreSQL instance and the local database.
Can I set the data synchronization mode between the primary/standby instance nodes?
The data synchronization method displayed on the RDS-PostgreSQL Console refers to the data synchronization method between the primary instance and the standby instance. Synchronous, semi-synchronous, and asynchronous methods are supported. The differences between the three synchronization methods are listed below:
• Synchronous: Verify that data changes for a transaction are synchronized to at least one Slave to improve data reliability.
• Semi-synchronous: Between asynchronous replication and synchronous replication. The system automatically increases or decreases the replication level according to the cluster status.
• Asynchronous: When data is modified, the Master returns data to the client immediately after execution, usually with high performance.
What is the time delay for primary/standby synchronization?
If the replication thread is running normally, the replication delay is a value representing seconds. The delay for primary/standby replication cannot be calculated using a formula because the delay is affected by the following factors:
• The transaction workload of the primary instance, that is, the time for a transaction to execute.
• Machine load status of the primary/standby instances.
You can view delay information in the RDS-PostgreSQL console dashboard. For details, see Dashboard.
Can multiple ECSs use the same RDS-PostgreSQL database?
In terms of usage, RDS-PostgreSQL has no differences from the PostgreSQL database built by you. Therefore, multiple ECSs can use the same RDS-PostgreSQL to support the business as long as the capability limits of a database are not exceeded.
Can I encrypt the disk after purchasing an RDS-PostgreSQL instance?
RDS-PostgreSQL does not support disk encryption. To encrypt data storage, you can use the transparent encryption function. For details, see Data Encryption.
Will different RDS-PostgreSQL instances share CPU and memory?
The CPU and memory refer to the purchased instance specifications. The resources between different instances are isolated and independent of each other without mutual affecting. Different RDS-PostgreSQL instances cannot share resources.
How many databases can run on an RDS-PostgreSQL instance?
The number of databases that can run on an RDS-PostgreSQL instance depends on your parameter settings, resource limits, etc.
When the resources (CPU, memory, disk, etc.), there are no limitations to the number of databases and the number of instances running on an RDS-PostgreSQL instance. However, if the number of database tables and data is too great, the instance backup rate will decrease.
1.1.2. Instance Freezing/De-registration/Renewal/
Unsubscription
Why is my RDS-PostgreSQL instance released?
After you purchase the eSurfing Cloud product, if your account is in arrears and you fail to renew or top up your account in time, the instance will be frozen and enter the retention period. If your account is still not renewed or topped up after the retention period expires, the instance enters the de-registration process. The data stored in the cloud service will be deleted and the cloud service resources will be released.
Why is my RDS-PostgreSQL instance frozen?
Your instance may be frozen for several reasons. The most common reason is that your account is in arrears or expires.
Can I still back up data if my RDS-PostgreSQL instance is frozen?
No. You cannot continue to back up data if your instance is frozen. If your instance is frozen due to arrears, you need to renew and unfreeze your instance first before backup.
How do I unfreeze a frozen database instance?
You can unfreeze your resources by renewal or topping up and then restore the normal use of your database instance. Database instances frozen due to arrears can be renewed, de-registered, or unsubscribed.
What happens when my instance is frozen, unfrozen, or de-registered?
It depends on different situations:
• After your instance is frozen: It cannot be accessed and used, causing interruption of ongoing business. In addition, the database instance cannot be connected to. When your database instance is frozen, no changes can be made. However, you can unsubscribe from the instance manually.
• After your instance is unfrozen: The restriction is released, and you can re-reconnect the instance to the database resource.
• After your instance is de-registered: The instance resources will be recycled and the instance will be deleted.
How do I renew my database instance?
After a yearly/monthly database instance expires, you can renew it on the Instance Management page of the Console. For detailed operations, see Product Renewing and Unsubscribing in the Billing Instructions.
Can I restore my instance if it is unsubscribed?
After your instance is unsubscribed, the underlying resources are recycled, and all user data cannot be restored. For this reason, make sure that you have backed up your data and confirmed that you do not need this instance and related data before unsubscribing from it.
Can I retrieve my instance if it is unsubscribed from by mistake?
No. Make sure that you have confirmed the information related to the instance before unsubscribing from it.
Why is my RDS-PostgreSQL instance still billed after being unsubscribed from?
After an RDS-PostgreSQL instance is unsubscribed from, the virtual machine (VM) used will stop billing. However, other resources, including Elastic IP (EIP), are still billed.
How do I unsubscribe from my RDS-PostgreSQL Instance?
You can unsubscribe from your database instance on the Console. For details, see Product Renewing and Unsubscribing in Billing Instructions.
1.1.3. Resources and Disk Management
How long does it take to create an instance?
Normally, it takes about 7-9 minutes to create an RDS-PostgreSQL single instance or primary/standby instance. If it takes more than 20 minutes, an error may occur during the creation. In this case, please contact your customer service staff.
Which types of logs and files occupy RDS-PostgreSQL disk space?
The following types of logs and files of the RDS-PostgreSQL instance occupy the disk space:
• File type Log files: Database log files, proxy log files, and high availability log files.
• Database files: Database content files.
Does RDS-PostgreSQL support scaling down the disk space?
For data security purposes, RDS-PostgreSQL does not support scaling down the disk space. You can only extend the disk space. We recommend that you set an appropriate disk space according to your business plan when enabling an instance.
Which items occupy the storage space of my RDS-PostgreSQL instance?
The items that occupy the storage space of your RDS-PostgreSQL instance include your regular data (backup files not included), the data required for the operation of your database instance, the log files generated by the database server, proxy log files, and the high-availability files (the proxy log files and the high-availability log files are cleaned up regularly according to the default policy).
How much disk space is required for DDL operations?
In general, DDL (Data Definition Language) operations do not have special disk space requirements on the disk space. However, some specific DDL operations require sufficient free disk space to handle. This depends on the specific DDL operation. To ensure the normal operation of your database, we recommend that you avoid DDL operations or other operations that can cause sharp increase of disk space occupation during peak hours.
1.1.4. Database Storage
What types of storage does RDS use?
Both RDS-PostgreSQL data storage and backup storage use EVS for storage.
For backup storage, the new storage space is used without occupying your database space.
Data storage supports three storage types: ordinary IO, high IO, and ultra-high IO. Currently, backup storage only supports ordinary IO storage types.
How do I view the RDS storage usage?
1. Log in to the RDS-PostgreSQL Console page.
2. Click the dashboard in the menu bar on the left and select the instance you want to view.
3. You can view the size and usage of the storage space of the instance in the disk capacity of the Information Display part.