View here to log in or access your console


Table Store

A fully managed NoSQL cloud database service that enables storage of massive amount of structured and semi-structured data

Buy Now Contact Sales


Alibaba Cloud Table Store is a scalable and fully managed NoSQL database service based on automatic data partitioning and load balancing technologies. Based on SSD technology, this cloud NoSQL database service enables you to store large quantities of structured and semi-structured data with real-time access, strong consistency and single-digit millisecond latency.

Table Store is an ideal fit for applications that require high memory and throughput such as IoT, games and mobile applications.


“The elasticity and scalability provided by Alibaba Cloud ECS ensures that our IT infrastructure can flexibly expand alongside our business growth. What’s more, Alibaba Cloud Table Store offers a convenient and flexible storage solution to efficiently handle massive volumes of data.”


Highly Available

  • Automated failure detection and quick failure recovery.

  • Guaranteed service availability of 99.9%.


  • Multiple cloud data backups and provides quick recovery in case of backup failure.

  • Guaranteed service reliability of 99.9999999%.


  • Automatic partitioning and SSD based technologies.

  • Offers processing capacity of over 10,000 queries per second for each node in the cluster.

  • Provides seamless scaling of applications hosted on the cloud.


  • Easily scales reserved resources based on real-time application hosting needs.


  • Authenticates each request to prevent unauthorized data access to NoSQL servers.

  • Offers user-level data isolation, access control and permission management.

  • Built-in technologies to mitigate DDoS and CC attacks on NoSQL servers.


  • Allows reserve resources for each table based on estimated data throughput.

  • Easily scales reserved resources based on an applications’ real-time needs

Easy to Use

  • Offers RESTful API, web-based Management Console and SDKs for multiple programming languages.

  • Complete product documentation such as Developer Guide, API references for developers.

Product Details

Built on Alibaba Cloud’s Apsara distributed system, Table Store organizes data in tables consisting of rows and columns, which are capable of scaling rapidly and seamlessly by leveraging data partitioning and load balancing technologies.

In regards to security, Table Store offers user-level data isolation, access control and permission management, as well as built-in technologies to mitigate DDoS and CC attacks on NoSQL servers. Table Store runs automated backups and provides quick recovery in case of backup failure.


Fully Managed

  • Automatically manages data partitions, software and hardware upgrades, configuration and cluster scaling out.

  • Relieves developers from day-to-day operational task of managing a NoSQL database.


  • Delivers consistent and fast performance for multiple application hosting on large-scale servers.

  • Provides stability for any row-level data operation.

Flexible Data Model

  • Supports applications to store data in multiple data types including integer, double, string, binary and Boolean.

  • Allows each data row to include irregular number of columns without predefining table schema.


  • Easy-to-use Management Console to monitor real-time usage of resources, including information such as requests per second and average response latency.

  • Provides detailed table-level monitoring information including data size, QPS, read/write capacity of units and average NoSQL server-side response time.

Multiple Operations Support

Table Operations

Provides table operations such as ListTable, CreateTable, and DescribeTable.

Data Operations

Performs data operations on a single row, multiple rows or even a particular range of rows.

Data Writing Operations

  • Supports operations such as PutRow, UpdateRow and DeleteRow, which guarantee atomicity and strong consistency.

  • Updates data in distributed file system automatically when data writing operation is successful and fetches latest data from written row.

Data Types of Column Values

Table Store supports 5 data types of column values:

Data TypeDefinitionIf PK Column SupportsSize Limitation
StringUTF-8, could be emptyYESNot greater than 2 MB and not greater than 1KB as primary key column
Integer64 bit IntegerYES8 Bytes
Double64 bit DoubleNO8 Bytes
BooleanTrue/FalseNO1 Byte
BinaryCould be NullYESNot greater than 2 MB and not greater than 1KB as primary key column


Table Store uses the following four dimensions to meter the resources used by applications and generate the bills: Data storage, the reserved read/write throughput, the amount of data transferred, the additional read/write throughput and the additional read/write throughput.

The following prices are for reference only. The exact price will be based on the customized plan you choose.

Choose your plan

Data Storage

Table Store charges hourly fees based on the total volume of instance data. Due to constant changes in the total data volume, Table Store collects statistics on the total data volume of all table partitions at regular intervals to calculate the average total data volume per hour. The average value is multiplied by the unit price to get the billing fee.

Data Storage
RegionChina MainlandHong KongSingaporeUS West/US EastJapan/Frankfurt/ Sydney/JakartaDubaiKuala Lumpur/ Mumbai
Capacity instance(Hourly subscription)US$0.00010/GBUS$0.00012/GBPlaningPlaningUS$0.00012/GBUS$0.00015/GBUS$0.00012/GB
High-performance instance(Hourly subscription)US$0.00030/GBPlaningUS$0.00030/GBUS$0.00030/GBPlaningPlaningPlaning

Data Transfer to Internet

Table Store charges fees when applications access the Internet downstream traffic of Table Store. Applications’ use of the HTTP method to access the responses returned by Table Store is the main component of the downstream traffic. Even if the operation fails, the operation failure information returned by Table Store will still produce downstream traffic.

Table Store only charges for the Internet downstream traffic, not for the intranet downstream traffic or the Internet upstream traffic. The access among different regions also belongs to the Internet access.

RegionChina MainlandHong KongSingaporeUS West/US EastJapanFrankfurt
PricingUS$0.123/GBUS$0.153/GBUS$0.081/GBUS$0.076/GBUS$0.120/GBUS$ 0.070/GB
RegionDubaiSydneyKuala LumpurMumbaiJakarta

Additional Throughput

The additional throughput is the portion of the actual consumed read/write throughput that exceeds the reserved read/write throughput per second. The additional read/write throughput is measured every second.

Table Store accumulates the additional read throughputs and write throughputs of all tables in an instance during every billing cycle. The actual consumed additional throughput is multiplied by the corresponding unit price to get the billing fee.

The Additional Read Throughput
RegionChina MainlandHong KongSingaporeUS West/US EastJapan/Frankfurt/Dubai/ Sydney/JakartaKuala Lumpur/ Mumbai
Capacity InstanceUS$0.0006/10K CUUS$0.0006/10K CUPlaningPlaningUS$0.0006/10K CUUS$0.00057/10K CU
High-performance InstanceUS$0.0030/10K CUPlaningUS$0.0030/10K CUUS$0.0030/10K CUPlaningPlaning
The Additional Write Throughput
RegionChina MainlandHong KongSingaporeUS West/US EastJapan/Frankfurt/Dubai/ Sydney/JakartaKuala Lumpur/ Mumbai
Capacity InstanceUS$0.0012/10K CUUS$0.0012/10K CUPlaningPlaningUS$0.0012/10K CUUS$0.00114/10K CU
High-performance InstanceUS$0.0060/10K CUPlaningUS$0.0060/10K CUUS$0.0060/10K CUPlaningPlaning

Reserved Throughput

The reserved read/write throughput is a table’s attribute. You can set a proper reserved read/write throughput for your data tables to reduce the costs of resource usage.

Table Store charges an hourly fee for the total reserved read/write throughput of all tables in an instance. The configured reserved read/write throughput may change constantly. Table Store collects the tables’ reserved read/write throughput at regular intervals to calculate the hourly average throughput. The average value is multiplied by the unit price to get the billing fee.

The Reserved Read Throughput
RegionChina MainlandHong KongSingaporeUS West/US EastJapan/Frankfurt/Dubai/ Sydney/JakartaKuala Lumpur/Mumbai
Capacity Instance(Hourly subscription)Not SupportedNot SupportedNot SupportedNot SupportedNot SupportedNot Supported
High-performance Instance(Hourly subscription)US$0.0001/CUPlaningUS$0.0002/CUUS$0.0001/CUPlaningPlaning
The Reserved Write Throughput
RegionChina MainlandHong KongSingaporeUS West/US EastJapan/Frankfurt/Dubai/ Sydney/JakartaKuala Lumpur/ Mumbai
Capacity Instance(Hourly subscription)Not SupportedNot SupportedNot SupportedNot SupportedNot SupportedNot Supported
High-performance Instance(Hourly subscription)US$0.0003/CUPlaningUS$0.0003/CUUS$0.0003/CUPlaningPlaning


Storing Internet-based User Information

Storing Internet-based user information and a NoSQL database service is a typical application scenario of the Alibaba Cloud Table Store. In such scenarios, Table Store saves structured/semi-structured data of end users—such as emails, journals, agenda and user information—in the Table Store for highly reliable storage that can be accessed anytime, anywhere.


  • Alibaba Cloud Table Store ensures consistent query performance even as data volume and access frequency increases.

  • Alibaba Cloud Table Store is already available for such use cases. For example, Alibaba Cloud Email, Cloud OS and Cloud Space. Such applications read/write the Table Store evenly and the data distribution is also even.

Storing Large Size Metadata

Alibaba Cloud Table Store can be utilized for storing metadata requirements running into hundreds of terabytes. Each piece of data includes multi-dimensional attributes which may increase or decrease any time. Therefore, it is hard to define a strict data model.

The number of page views can also escalate into high numbers. Queries can run into trillions per day peaking at 100,000 QPS, and requires latency in milliseconds.

Alibaba Cloud Table Store is ideal for such scenarios as it enables consistent read/write, wherein new records and updates can be made. Today many applications are using Alibaba Cloud Table Store, such as Alibaba Cloud MaxCompute metadata (MaxCompute table/Job/Security and related metadata are stored in Alibaba Cloud Table Store).

Storing Logs and Monitoring Data

Such applications when hosting typically experience a steady and large inflow of data and require persistent storage of hot data. The application normally keeps the latest data and updates to the most recent period of time (e.g. 1 month or half a year) and discards older data.

Applications are required to store and manage logs, as well as monitor data. Such applications write Table Store in batches and the read volume is not large. Alibaba Cloud Log Service is a typical example of this use case.

Getting Started

Get access to Table Store CLI, APIs and SDKs to manage your databases. Below are the required links:

Using Table Store through Management Console

The Alibaba Cloud Management Console provides a simple web-based user interface which is used to access and configure the Table Store resources. Using this console, you can create and modify your NoSQL hosting databases, change the capacity of the resources being used and take various backups.

Using Table Store With the Alibaba Cloud Table Store CLI

Alibaba Cloud Table Store gives you the freedom to use the Table Store resources and manage them through the Command Line Interface (CLI). You can download and install the Table Store CLI package to enable this service.

Table Store API Reference

You can use web based Table Store APIs to create or modify Databases. You can control access, secure your servers or Instances and create several backups through this service. Here is the Alibaba Cloud Table Store API Guide with a full list of the available APIs.


Alibaba Cloud Table Store is a highly scalable service, based on automatic data partitioning and load balancing technologies. It is an ideal fit for application hosting which requires high memory and throughput such as IoT, games and mobile applications.

Leverage the benefits of Table Store using the Alibaba Cloud Management Console and Alibaba Cloud documentation.

Below are the links to the documentation, SDKs, and other resources.

Developer Resources


1. Is Table Store a database? What is the difference between Table Store and the conventional Relational Database Service (RDS)?

Alibaba Cloud Table Store is a NoSQL managed data storage service. The cloud based NoSQL database service has a distributed structured and semi-structured data model. This is different in comparison to a conventional RDS database that supports MySQL and SQL Server.

  • The data model of Table Store is a 2D-table, centered with rows and columns. But unlike a conventional database, the table of Table Store is sporadic with different columns allowed in one row.

  • In the Table Store, the attribute columns can be dynamically added or reduced and no strict schema is required for the attribute columns at the time of table creation.

  • Compared to the rich functions of traditional databases such as views, indexes, transactions and abundant SQL statements, Table Store offers relatively basic functions with better scalability. This makes it easier to support a large volume of data such as hundreds of TB and concurrent queries (100,000 QPS for a single table).

  • In programming, Table Store offers uniform HTTP Restful APIs and does not support traditional SQL conventions. It is flexible and allows you to pay for the actual usage resources used for storage and read/write throughput.

2. How is data secured on Alibaba Cloud Table Store?

Table Store ensures data security in two ways:

  • Security of Data Storage (data reliability): The Table Store guarantees 99.9999999% data reliability. When data is stored on the bottom layer of Table Store, multiple copies are created. When one copy of data has issues like data loss or data corruption, the repair procedure is automatically triggered to quickly replace the missing data. This ensures no data loss from Table Store.

  • Security of Data Access (strict identity authentication for access): Every user needs to create the AccessKeyID and AccessKeySecret. The AccessKeySecret is confidential to the user. The user must provide the correct AccessKeyID on every query request as the signature and Table Store will verify the AccessKeySecrect for every query request. The follow-up operation is allowed only after verification.

3. Is the attribute column a requisite during table creation?

No. Table Store supports semi-structured tables. Only the primary key columns (Column 1-4) are required for table creation and the attribute columns are not required. There can be multiple attribute columns in Table Store (The number of columns for each row has no limit) and each row of data can have a different number of attribute columns. When the application is writing data, Table Store requires it specify the names and values of all the data columns (primary key columns and attribute columns).

4. How can I understand the partition key in the first column of the primary key during table creation?

When the table data size reaches a set value Table Store will partition the table, based on a range of values in the partition key column, in order to achieve load balancing.

At the time of table creation, the table has one partition by default—all data is on the same partition of the table. When the table has multiple partitions, data in each partition falls within a certain range of values in the partition key column. All values in the partition key column are segmented by the natural order of column values, by the natural order of Integer or String (data type in the primary key column).

In addition to query performance, partitioning will also affect the throughput rate of reserved read/write. When there are multiple partitions in the table, reserved read/write throughput will be proportioned to each partition.

5. How do I set a reasonable partition key?

The selection of the partition key is very important when creating tables because it will affect query performance if the data size is too large. When selecting the partition key for applications, users can follow basic principles as below:

  • Do not use attributes with a fixed value or a small value range, such as user gender (Male/Female).

  • Avoid attributes that will have obvious query hotspots after sorting by natural order, such as using TimeStamp as the partition key to query the latest data scenarios.

  • Use attributes with decentralized query hotspots after sorting by the natural order, such as UserID.

6. What if I cannot predict and control the query hotspot of the application?

We recommend to hash the data before introducing a partition key according to the application characteristics. For example, when writing a row of data, generate a hash value for the UserID using the simple hash algorithms. Then splice the hash value and the UserID and save them as the partition key value into the table. This lightweight measure can effectively solve query hotspot issues. But remember, since the partition key value is a result of spliced hash value and the actual value, the application will not be able to read the range (getRange) with the partition key.

7. What is the limit on the number of tables under an account?

Each Table Store user can create ten instances at most and each instance can have 64 tables at most. Therefore each Table Store user can create a maximum of 100 tables under one account.

8. What if the data size is large and query performance requirement is high?

Unlike conventional SQL databases (MySQL/SQL Server) which address massive data access demands by database sharding and table partitioning, Table Store adopts the distributed approach and cracks the bottleneck of huge data size and access latency. Users can save structured or semi-structured data in a sparse table (sparse table is a data structure which often serves as a substitute for segment tree in case of immutable data), without worrying about query performance in the case of huge data size.

Due to the distributed feature of Table Store, the number of tables also become a resource attribute. Although the number of tables have a maximum value under one account for service resource control, the scalability of Table Store can effectively address the table count limitation.

If you still need to increase the quota of tables under one account, you can open a ticket or contact your Alibaba Cloud customer manager.