| 2018-12-20 |
Java UDF development through Intellij IDEA |
Experience optimization |
You can use Intellij IDEA to develop Java UDFs. |
Use IntelliJ IDEA to develop a Java-based UDF |
| 2018-12-04 |
MaxCompute for EU West 1 (London) |
New region/zone |
MaxCompute is rolled out to provision services in EU West 1 (London). You can purchase
resources and activate projects as needed on the Alibaba Cloud console of the Intl,
China, or Japan edition.
|
Configure endpoints |
| 2018-11-26 |
JSON data extraction from MongoDB to MaxCompute |
Experience optimization |
You can use the data synchronization feature of DataWorks to extract JSON data from
MongoDB to MaxCompute.
|
Migrate JSON data from MongoDB to MaxCompute |
| 2018-11-15 |
External tables |
Experience optimization |
The topic that describes how to process unstructured data is replaced with the "Overview
of external tables" topic.
|
Overview of External tables |
| 2018-11-13 |
Migration of JSON data from OSS to MaxCompute |
Experience optimization |
You can use the data synchronization feature of DataWorks to migrate JSON data from
OSS to MaxCompute. In addition, you can use the GET_JSON_OBJECT function provided by MaxCompute to extract JSON data.
|
Migrate JSON data from OSS to MaxCompute |
| 2018-11-08 |
MaxCompute limits |
Experience optimization |
The original "Reading guidance" topic is replaced with the "What you must know" topic.
The new topic contains limits of MaxCompute SQL, data upload and download, MapReduce,
security configuration, lighting, and PyODPS, which are not provided in the original
topic.
|
What you must know |
| 2018-11-05 |
cume_dist window function of MaxCompute SQL
|
New feature |
You can use the cume_dist window function to calculate the accumulative distribution of data in a group. That
is, the proportion of rows with values that are smaller than or equal to the current
value in all rows in the group.
|
Window functions |
| 2018-11-02 |
Dynamic partitioning based on data migration from RDS to MaxCompute |
Experience optimization |
You can use the data synchronization feature of DataWorks to automatically create
partitions and dynamically migrate data from RDS to MaxCompute.
|
Migrate data from RDS to MaxCompute to implement dynamic partitioning |
| 2018-09-29 |
User Defined Type (UDT) |
New feature |
You can reference third-party classes or objects in SQL scripts to obtain the data
or call the methods of these classes or objects.
|
UDT |
| 2018-09-29 |
User Defined Join (UDJ) |
New feature |
UDJ is an extension to the UDF framework of MaxCompute. With UDJ, you can define join
operations and perform these join operations across tables flexibly. In addition,
UDJ reduces the need for performing operations at the bottom layer of the distributed
system, making the system run faster.
|
UDJ |
| 2018-07-30 |
Solutions to long-tail traffic |
Experience optimization |
The documentation is updated to describe typical long-tail traffic issues and the
corresponding solutions.
|
Optimize long tail computing |
| 2018-05-25 |
GZIP compression |
New feature |
MaxCompute can use its built-in extractor to read CSV or TSV data that is compressed
by using GZIP from OSS. An uncompressed file differs from a compressed file in their
SERDEPROPERTIES attributes.
|
Access OSS unstructured data |
| 2018-05-25 |
Hive |
New specification |
With Hive, you can use the unstructured data framework of MaxCompute to process the
data that is stored on OSS in a variety of open-source formats such as ORC, PARQUET,
SEQUENCEFILE, RCFILE, AVRO, and TEXTFILE.
|
Processing open source format data for OSS |
| 2018-05-25 |
SELECT TRANSFORM statement
|
New feature |
The SELECT TRANSFORM statement simplifies code reference and script compilation. It supports a wide range
of programming languages such as Java, Python, Shell, and Perl, helping you to develop
ad hoc functions.
|
SELECT TRANSFORM |
| 2018-05-25 |
SEMI JOIN statement
|
New feature |
In a SEMI JOIN statement, the right-side table is used only to filter the data from the left-side
table and will not be presented in the result set. This statement supports two clauses:
LEFT SEMI JOIN and LEFT ANTI JOIN.
|
SEMI JOIN |
| 2018-05-25 |
Implicit type conversion in CASE WHEN expressions
|
New specification |
MaxCompute SQL allows for implicit type conversion in CASE WHEN expressions.
|
Type conversions |
| 2018-05-16 |
Full table scan |
New feature |
You can enable and set the full table scan function. |
Other operations |
| 2018-04-14 |
Renewal of expired instances through Auto-Renew |
New specification |
When your Subscription instance expires, you can renew it through the Auto-Renew option on the MaxCompute console.
|
None |
| 2018-04-10 |
MaxCompute for US East 1 (Virginia) |
New region/zone |
MaxCompute is rolled out to provision services in US East 1 (Virginia). You can purchase
resources and activate projects as needed on the Alibaba Cloud console of the Intl,
China, or Japan edition.
|
Configure endpoints |
| 2018-04-02 |
MaxCompute Studio 2.9.0 |
New feature |
MaxCompute Studio 2.9.0 enables you to check and rectify code, develop and debug graphs,
and fix errors. For more information, see MaxCompute Studio 2.9.0.
|
What is Studio |
| 2018-03-28 |
MaxCompute Manager for China East 1 (Hangzhou) |
New specification |
Enterprise customers who use the Subscription billing method in China East 1 (Hangzhou)
can get their computing resource usage and task queuing information. In addition,
MaxCompute Manager can manage computing tasks by group based on task priorities.
|
MaxCompute CU Management |
| 2018-03-23 |
Tunnel endpoint customization in MaxCompute Studio |
Defect rectification |
When you use Instance Tunnel to download SQL query results, you can specify a tunnel
endpoint.
|
Configure endpoints |
| 2018-03-22 |
MaxCompute for China East 1 (Hangzhou) |
New region/zone |
MaxCompute is rolled out in China East 1 (Hangzhou) to provision services. You can
purchase resources and activate projects as needed on the Alibaba Cloud console of
the Intl, China, or Japan edition.
|
Configure endpoints |
| 2018-02-09 |
PyODPS 0.7.15 |
New version/New specification |
The relevant features in PyODPS 0.7.15 are optimized or rectified and new features
are introduced to enhance development experience and efficiency:
- You can use DataFrame to define functions based on Python 3.6.
- The defect due to which the functions that you define through DataFrame in Python
3.5 incur errors is fixed.
- You can set the parameters that determine the running of XFlow when XFlow is operating.
- When you create a PyODPS DataFrame from a Pandas DataFrame, you can use the as_type parameter to specify the combination type, for example, LIST or DICT.
|
Python SDK |
| 2018-02-09 |
MaxCompute for Asia Pacific SE 5 (Jakarta) |
New region/zone |
MaxCompute is rolled out to provision services in Asia Pacific SE 5 (Jakarta). You
can purchase resources and activate projects as needed on the Alibaba Cloud console
of the Intl, China, or Japan edition.
|
Configure endpoints |
| 2018-02-03 |
MaxCompute Studio 2.8.2.3 |
Experience optimization |
- In the left-side pane of the MaxCompute Studio Editor, you can directly run an SQL
statement without clicking any button in the toolbar.
- In the MaxCompute Studio Project Explorer, built-in functions are classified by type
to help you easily find a specific function.
|
None |
| 2018-02-03 |
top instance command
|
New feature |
The documentation is updated to describe the top instance command.
|
Instances operations |
| 2018-02-03 |
external info query
|
New feature |
The documentation is updated to describe the external info element, including LOCATION and StorageHandler, of external tables.
|
Table operations and Access OSS unstructured data |
| 2018-02-01 |
top instance command
|
New specification |
In MaxCompute Console 0.29.0, you can run the top instance command to obtain information about the jobs that running or waiting in the queue.
|
Instances operations |
| 2018-01-31 |
MaxCompute Manager |
New feature |
Enterprise customers who use the Subscription billing method can get their computing
resource usage and task queuing information. In addition, MaxCompute Manager can manage
computing tasks by group based on task priorities.
|
MaxCompute CU Management |
| 2018-01-22 |
DataHub for four regions |
New region/zone |
DataHub is introduced to four regions: China East 2 (Shanghai), China North 2 (Beijing),
China South 1 (Shenzhen), and Asia Pacific SE 1 (Singapore).
|
None |
| 2018-01-18 |
MaxCompute Studio 2.8.2.2 |
New version/New specification |
MaxCompute Studio 2.8.2.2 provides the following functions:
- Develop Python UDFs.
- Compile unstructured unit tests.
- Operate complex data types while running UDFs on your computer.
- Define SQL statement format rules.
|
What is Studio |
| 2018-01-18 |
MaxCompute for Asia Pacific SOU 1 (Mumbai) |
New region/zone |
MaxCompute is rolled out in Asia Pacific SOU 1 (Mumbai) to provision services. You
can purchase resources and activate projects as needed on the Alibaba Cloud console
of the Intl, China, or Japan edition.
|
Configure endpoints |
| 2018-01-10 |
Full table scan |
New feature |
The full table scan function is forbidden by MaxCompute so as to avoid resource waste
and reduce costs.
|
Project operations |
| 2018-01 |
Project data protection |
Experience optimization |
The documentation is updated to describe the background information and mechanism
of data protection as well as how to allow for data outflow when the data protection
mechanism is enabled.
|
Project data protection |
| 2018-01 |
Hive UDF |
Experience optimization |
An example is added to describe how to use Hive UDFs in MaxCompute. |
Java UDF |
| 2018-01 |
TableTunnel |
Experience optimization |
The documentation is updated to describe TableTunnel, its limits, and how it works
at the bottom layer.
|
TableTunnel |
| 2018-01 |
Full table scan |
New feature |
The documentation is updated to describe the full table scan function. |
Announcements, Other operations, and SQL Common Errors |