ApsaraDB RDS Glossary

更新时间:
复制 MD 格式

Quick reference for terms used in ApsaraDB RDS.

A

  • AliPG AliPG is the database kernel that powers Alibaba Cloud's PostgreSQL-compatible cloud database services. Developed by Alibaba Cloud and in commercial use since 2015, AliPG is compatible with open source PostgreSQL and supports all mainstream PostgreSQL versions. It has operated stably across internal Alibaba Group services and Alibaba Cloud since launch. For more information, see AliPG benefits.

  • Asynchronous mode In asynchronous mode, the primary RDS instance returns a success response to the application immediately after completing a write operation, without waiting for the secondary instance to confirm replication. The primary instance replicates the change to the secondary in parallel. If the secondary becomes unavailable, workloads on the primary continue uninterrupted. However, if the primary fails before replication completes, a small amount of data may not have reached the secondary, causing data inconsistencies between the primary and secondary instances. Contrast with semi-synchronous mode and synchronous mode.

B

C

  • Character set A character set is a collection of letters, symbols, and encoding rules used to represent and store text in a database.

  • Classic network In the classic network, RDS instances share a flat network with other resources and cannot be isolated at the network level. Access control relies entirely on IP address whitelists or security groups. For better isolation and security, use a VPC instead.

  • CPU utilization CPU utilization measures the percentage of CPU capacity consumed by database processes over a given period.

D

  • DAS Database Autonomy Service (DAS) is an Alibaba Cloud service that uses machine learning and expert knowledge to automate database perception, self-healing, optimization, O&M, and security. DAS reduces the risk of service disruptions caused by manual errors and keeps databases running stably, securely, and efficiently. For more information, see Overview of DAS.

  • Database A database is a logical container within an RDS instance that stores data with minimal redundancy. Data in a database is independent of application code and can be shared across multiple authorized users. Each RDS instance supports multiple databases, and each database name must be unique within that instance.

  • Deadlock A deadlock occurs when two or more concurrent transactions each hold a lock on a resource that the other transaction needs, and neither can proceed. Most deadlocks arise when multiple transactions try to modify the same set of resources in conflicting orders.

  • Dedicated RDS instance A dedicated RDS instance has exclusively allocated CPU and memory resources, so its performance is never affected by other instances on the same physical host. The highest-tier configuration within this family is the dedicated host type, which occupies all resources on a single physical machine. For more information, see Instance families.

  • DMS Data Management (DMS) is a GUI-based tool developed by Alibaba Cloud for managing databases and servers. DMS supports data management, schema management, server management, access control, business intelligence (BI) charting, data trend analysis, data tracking, and performance monitoring and optimization. It works with relational databases (MySQL, SQL Server, PostgreSQL, MongoDB, Redis), NoSQL databases, and Linux-based database servers. For more information, see What is DMS?

  • DTS Data Transmission Service (DTS) is an Alibaba Cloud data streaming service for migrating, subscribing to, and synchronizing data between relational databases, NoSQL databases, and OLAP databases over stable, secure links. For more information, see What is DTS?

E

  • ECS Elastic Compute Service (ECS) is a cost-effective, easy-to-use, easy-to-maintain, secure, reliable, and scalable service provided by Alibaba Cloud. ECS instances support a wide range of workloads including websites, applications, and batch jobs. ECS is commonly paired with ApsaraDB RDS to build application backends. For more information, see What is ECS?

  • ESSD An Enhanced SSD (ESSD) is a cloud disk developed by Alibaba Cloud on a next-generation distributed block storage architecture. ESSDs use 25 Gigabit Ethernet and Remote Direct Memory Access (RDMA) to deliver ultra-low latency and up to 1 million random read/write IOPS. Three performance levels are available: For more information, see Storage types.

    • PL1 ESSD: The baseline performance level.

    • PL2 ESSD: Delivers approximately 2x the IOPS and throughput of a PL1 ESSD.

    • PL3 ESSD: Delivers up to 20x the IOPS and 11x the throughput of a PL1 ESSD. Best suited for workloads requiring high concurrency, high I/O performance, and low read/write latency.

F

  • Full data migration Full data migration copies all data from a source RDS instance to a destination RDS instance in a single pass. Use DTS to run a full data migration. After it completes, run incremental data migration to synchronize changes that occurred during migration.

I

  • I/O Read and write operations performed on a disk.

  • IOPS Input/output operations per second (IOPS) measures the number of 4 KB read or write operations a block storage device can handle per second.

  • Incremental data migration After full data migration completes, incremental data migration uses DTS to continuously synchronize new changes (INSERT, DELETE, and UPDATE operations) from the source instance to the destination. This keeps the destination in sync and allows you to cut over with minimal downtime.

  • IP address whitelist An IP address whitelist defines the IP addresses and CIDR blocks that are allowed to connect to an RDS instance. By default, only the loopback address 127.0.0.1 is included. Update the whitelist regularly to reflect changes in your network topology. For more information, see Configure an IP address whitelist.

J

  • JDBC Java Database Connectivity (JDBC) is a standard Java API for connecting to relational databases, including ApsaraDB RDS instances.

L

  • Local disk A local disk (local SSD) is storage co-located on the same physical host as the RDS instance. Because data travels a shorter physical path, local disks offer lower I/O latency than cloud disks. See also Premium Local SSD. For more information, see Storage types.

  • Logical I/O Logical I/O refers to read and write operations issued by an application to the file system layer, before the operating system determines whether to serve the request from cache or physical disk.

M

  • MaxCompute MaxCompute (formerly ODPS) is a fully managed, large-scale data computing platform that can process terabytes to petabytes of data at high speed. It supports comprehensive data import solutions and classic distributed computing models, making it a common destination for bulk data exports from ApsaraDB RDS. For more information, see What is MaxCompute?

  • Microsoft AD Microsoft Active Directory (AD) is a directory service developed by Microsoft for Windows Standard Server, Windows Enterprise Server, and Microsoft SQL Server. A directory is a hierarchical structure that stores information about network objects such as user accounts, passwords, and phone numbers, and makes that information available to authorized users on the same LAN. Many large enterprises use AD's domain control mechanism for centralized identity and access management. For more information, see Connect an ApsaraDB RDS for PostgreSQL instance to a self-managed AD domain.

  • MongoDB ApsaraDB for MongoDB is a stable, reliable, and scalable database service fully compatible with the MongoDB protocol. A common pattern is to store structured relational data in ApsaraDB RDS and unstructured or document data in ApsaraDB for MongoDB. For more information, see What is ApsaraDB for MongoDB?

O

  • OSS Object Storage Service (OSS) is a secure, cost-effective, and highly durable cloud storage service provided by Alibaba Cloud for storing large volumes of unstructured data. For more information, see What is OSS?

  • OpenAPI ApsaraDB RDS exposes a public API that lets you manage instances, databases, accounts, and other resources programmatically. For more information, see List of operations by function.

P

  • Pay-as-you-go Pay-as-you-go instances are billed per hour based on actual resource usage. This billing method suits short-term or variable workloads. Release the instance when it is no longer needed to stop incurring charges. For more information, see Billable items.

  • Performance insight Performance insight is a DAS feature that provides instance-level load monitoring, correlation analysis, and performance optimization. Use it to identify load bottlenecks and troubleshoot performance issues.

  • Premium Local SSD Premium Local SSD is a type of local SSD co-located on the same physical host as the RDS instance. Storing data on a Premium Local SSD reduces I/O latency compared to cloud-based storage. See also local disk. For more information, see Storage types.

Q

  • QPS Queries per second (QPS) is the number of SQL statements — including INSERT, SELECT, UPDATE, DELETE, and COMMIT — executed on an RDS instance per second.

R

  • Random I/O Random I/O refers to read and write operations that access non-contiguous locations on a disk. Contrast with sequential I/O.

  • RAM Resource Access Management (RAM) is an Alibaba Cloud service that manages user identities and resource access permissions. You can grant the permissions on specific RDS instances to RAM users. For more information, see What is RAM?

  • RDS ApsaraDB Relational Database Service (RDS) is a fully managed, scalable online database service built on the Apsara distributed file system and high-performance SSDs. It supports MySQL, SQL Server, PostgreSQL, and MariaDB engines and provides built-in solutions for disaster recovery, backup, restoration, monitoring, and migration. For more information, see What is ApsaraDB RDS?

  • RDS Basic Edition RDS Basic Edition is the entry-level edition with a single primary instance and compute-storage separation. It has no secondary standby instance, so it offers lower cost but no built-in high availability. For more information, see RDS Basic Edition.

  • RDS Cluster Edition RDS Cluster Edition for PostgreSQL is a high-availability architecture with one primary node and multiple secondary nodes, built on compute-storage separation. It provides automated failover, primary/secondary switchover, readable secondary nodes, node addition and deletion, multi-zone disaster recovery, node-level monitoring, and cluster topology management. For more information, see RDS Cluster Edition.

  • RDS High-availability Edition RDS High-availability Edition pairs each primary instance with a secondary standby. Data is replicated from primary to secondary in semi-synchronous mode. If the primary fails, the service automatically fails over to the secondary. For more information, see RDS High-availability Edition.

  • RDS instance An RDS instance is a virtual database server that exclusively occupies allocated memory resources on a physical host. Each instance can run a specific database engine with a defined memory size and storage capacity, and supports multiple databases.

  • Read-only RDS instance A read-only instance offloads SELECT traffic from the primary instance. Deploy one or more read-only instances when your application generates heavy read workloads, to increase overall read throughput without scaling the primary. For more information, see Overview of read-only ApsaraDB RDS for PostgreSQL instances.

  • Read-only QPS The number of SELECT statements executed per second on an instance operating in read-only mode.

  • Read and write QPS The number of INSERT, SELECT, and UPDATE statements executed per second on an instance handling both read and write traffic.

  • Recycle bin Expired or released RDS instances move to the recycle bin rather than being immediately destroyed. From the recycle bin, you can unlock, rebuild, or permanently destroy an instance. For more information, see Use the recycle bin feature.

  • Region A region is the geographic location where an RDS instance is deployed. Select the region closest to your users or downstream services at purchase time. The region cannot be changed after the instance is created.

S

  • Schema migration Schema migration uses DTS to copy the schema definitions of database objects (tables, indexes, views, and so on) from a source instance to a destination instance, without copying the data itself.

  • Semi-synchronous mode In semi-synchronous mode, the primary instance waits until at least one secondary instance acknowledges receipt of an update log record before committing the transaction. The data is considered safe as soon as one secondary confirms receipt — the log does not need to be replayed yet. If a secondary RDS instance is unavailable or a network interruption occurs between the primary and secondary RDS instances, the mode automatically degrades to asynchronous mode to keep the primary available. Contrast with synchronous mode, which requires all secondaries to confirm receipt and never degrades.

  • Sequential I/O Sequential I/O refers to read and write operations that access contiguous blocks on a disk in order. Contrast with random I/O.

  • SSL Secure Sockets Layer (SSL), developed by Netscape and later upgraded by the IETF to TLS, establishes encrypted connections between clients and servers. ApsaraDB RDS supports SSL for MySQL, SQL Server, and PostgreSQL engines. Each instance receives a server SSL certificate you can use to verify that the database service responding on a given IP address and port is genuinely your RDS instance — protecting against man-in-the-middle attacks. You can enable and renew the certificate at any time. For more information, see SSL encryption.

  • Standard SSD A standard SSD is an elastic block storage device built on a distributed storage architecture. Storing data on a standard SSD separates compute from storage, enabling independent scaling of each. For more information, see Storage types.

  • Subscription Subscription instances are billed upfront for a fixed term (monthly or yearly). Longer subscription terms come at lower effective rates. For long-running production workloads, subscription is more cost-effective than pay-as-you-go. For more information, see Billable items.

  • Synchronous mode In synchronous mode, the primary instance waits until at least one secondary instance acknowledges receipt of an update log record before committing the transaction. This provides strong durability guarantees but cannot degrade to asynchronous mode. Synchronous mode is available only in RDS Enterprise Edition, and the replication mode cannot be changed after the edition is selected. Contrast with semi-synchronous mode, which waits for at least one secondary and degrades automatically when secondaries are unavailable.

T

  • Tair Tair (Redis OSS-compatible) is a database service that provides a hybrid of memory and hard disks to ensure data persistence. Use Tair alongside ECS and ApsaraDB RDS to handle burst read traffic — for example, caching frequently accessed data to reduce load on the primary RDS instance. For more information, see What is Tair (Redis OSS-compatible)?

  • TDE Transparent data encryption (TDE) encrypts data files as they are written to disk and decrypts them as they are read into memory, with no changes required to your application or data file sizes. ApsaraDB RDS supports TDE for MySQL, PostgreSQL, and SQL Server engines. For more information, see Use the TDE feature.

  • TPM Transactions per minute (TPM) is the number of transactions executed on an RDS instance per minute.

  • TPS Transactions per second (TPS) is the number of transactions committed on an RDS instance per second.

V

  • VPC A virtual private cloud (VPC) is a logically isolated network environment within Alibaba Cloud, providing higher security and higher performance than the classic network. Instances in a VPC are isolated from other tenants at the network level. For more information, see What is a VPC?

Z

  • Zone A zone is a physically independent data center with its own power and network infrastructure within a region. All zones within a region offer equivalent services. Deploying resources across multiple zones within the same region provides protection against single-zone failures.