×
Community Blog Product Updates on Alibaba Cloud Products | November 2023

Product Updates on Alibaba Cloud Products | November 2023

In this article, we summarized the new features and experience enhancements of Alibaba Cloud Products in November 2023.

Alibaba Cloud International Product Updates showcase the latest news and information on our products, helping you to quickly discover and understand our latest product releases, updates, and offers on the Alibaba Cloud International Website.

New Features

ApsaraDB for Redis - Pay-as-you-go Billing Available for Bandwidth Upgrade

The pay-as-you-go billing method is available for the bandwidth upgrade feature. In this billing model, You are billed for this feature on an hourly basis.

Target customers:

  1. Businesses that require more bandwidth than what is offered
  2. Businesses that experience significant traffic fluctuations

Features released:

  1. The minimum billing duration for bandwidth upgrades has been changed from 1 day to 1 hour, resulting in a reduction in the associated costs.
  2. You can dynamically adjust the bandwidth of your instance based on the fluctuating traffic patterns of your business. The new billing model allows you to significantly reduce the costs associated with bandwidth upgrades while still ensuring that your bandwidth requirements are met.

ApsaraVideo Live - Support for GA Configuration

ApsaraVideo Live can take advantage of the acceleration capabilities of Global Accelerator (GA) to accelerate stream ingest and streaming.

Target customers:

Internet and gaming companies with a global business

Features released:

By leveraging the high-quality BGP bandwidth and global network infrastructure of Alibaba Cloud, GA allows users to connect to the nearest access points for content delivery acceleration. This helps improve user experience during live streaming and reduce network issues, such as network latency, network jitter, and packet loss.

ApsaraDB RDS for PostgreSQL - Releases Fully Encrypted Database (Basic Edition)

The fully encrypted database (basic edition) feature is supported for ApsaraDB RDS for PostgreSQL instances that use any instance types.

Target customers:

Customers who require high full-link security protection.

Features released:

ApsaraDB RDS for PostgreSQL allows you to use the index acceleration extension, more encrypted data types, more operators, and more clauses when you use fully encrypted database (basic edition) feature for your RDS instances that use instance types other than the Intel SGX-based security-enhanced instance types.

ApsaraDB RDS for PostgreSQL - Supports Cache Eviction

The cache eviction feature is supported for ApsaraDB RDS instances that run PostgreSQL 14 and PostgreSQL 15.

Target customers: users whose workloads are involved in software as a service (SaaS) scenarios with a large number of persistent connections.

Features released: The process-level cache eviction feature is supported to reduce the memory usage of persistent connections and ensure that your database does not run out of memory due to excessive persistent connections. The feature uses the Least Recently Used (LRU) policy to release caches that are infrequently used. This reduces the memory usage of connections and improves the stability of ApsaraDB RDS for PostgreSQL instances.

ApsaraDB RDS for PostgreSQL - Release notes of ApsaraDB RDS for PostgreSQL

ApsaraDB RDS for PostgreSQL provides the minor engine version of 20230830.

Target customers: all users.

Features released:

New features of the minor engine version of 20230830:

PostgreSQL 14 and PostgreSQL 15:

The cache eviction feature is supported.

The postgresql_anonymizer extension is updated to version 1.1.0.

The pgvector extension is updated to version 0.5.0 and supports Hierarchical Navigable Small World (HNSW) indexes, construction of IVFFLAT parallel indexes, element-by-element multiplication of vectors, L1 distance functions, and sum aggregation.

The extensions of Ganos are updated to version 5.8 for PostgreSQL 10, PostgreSQL 11, PostgreSQL 12, PostgreSQL 13, PostgreSQL 14, and PostgreSQL 15.

ApsaraDB RDS for MySQL - Releases Cost-effective Instance Types for RDS Cluster Edition

Cost-effective instance types for RDS Cluster Edition are more cost-efficient, which helps reduce costs and improve efficiency.

Target customers: users who want to reduce costs.

Features released: A total of 28 cost-effective instance types are released for RDS Cluster Edition of ApsaraDB RDS for MySQL. These instance types allow users to enjoy technical benefits and help reduce cloud resource costs.

ApsaraVideo Live - Support for HTTP/2 Configuration

Users can configure HTTP/2 by using the console or calling API operations.

Target customers: Intended for customers who use ApsaraVideo Live.

Features released: HTTP/2, originally named HTTP/2.0, is the first new version of HTTP since HTTP/1.1. HTTP/2 is a binary protocol that supports multiplexing and header compression. It greatly improves web performance and reduces network latency.

Platform for AI - DSW supports pay-as-you-go instances based on Elastic Container Instances and cloud disks

You can create instances based on Elastic Container Instances and cloud disks in DSW. The instances provide free storage and supports accelerated startups.

Target customers: all developers

Features released:

New types of pay-as-you-go DSW instances

  1. The instance startup speed is increased by over three times.
  2. The instance provides storage free of charge to persist the environment and data.

You can start and stop instances at any time based on your business requirements. This allows you to use resources in a flexible and efficient manner.

Platform for AI - EAS supports push-button deployment of the Tongyi Qianwen model

EAS supports push-button deployment of the Tongyi Qianwen model

Target customers: customers of model service and AI inference

Features released:

You can use PAI-EAS to deploy web UI applications that are based on the open source Tongyi Qianwen model, and use the web UI and API operations to perform model inference.

Qwen-7B is a 7 billion-parameter model of the Tongyi Qianwen series developed by Alibaba Cloud. Qwen-7B is a large language model based on Transformer, and is trained on ultra-large-scale pre-trained data. The pre-trained data covers a wide range of data types, including a large number of texts, professional books, and code. In addition, Qwen-7B-Chat, an LLM AI assistant is developed based on Qwen-7B.

Platform for AI - ChatGLM model deployment and LangChain supported in EAS

You can deploy ChatGLM applications and use LangChain in EAS.

Target customers: customers of model service and AI inference.

Features released:

Deploy the ChatGLM application and use Langchain to integrate business data in EAS.

ChatGLM is an open source dialogue language model developed based on the General Language Model (GLM) architecture and supports dialogues in both Chinese and English. ChatGLM-6B is optimized for dialogues in Chinese by using a technology similar to ChatGPT. ChatGLM-6B is trained on 1 trillion tokens, and uses optimization technologies such as supervised fine-tuning (SFT), self-feedback system, and human feedback reinforcement learning (RLHF) to better align with human preferences.

LangChain is an open source framework that allows AI developers to integrate large language models (LLM) and data from other sources. This way, you can use LLM with as little computing resources as possible.

Platform for AI - Automatic scaling of the asynchronous inference service is supported in EAS

The system automatically scales resources for the asynchronous inference service based on the queue length.

Target customers: customers of the model service and AI inference.

Features released: The EAS asynchronous inference service can automatically scale resources by monitoring the amount of data in the queue. This can help you manage the number of service instances in an effective manner, increase resource utilization, and achieve cost-effectiveness.

Platform for AI - Internet access through private NAT gateway is supported for DSW instances

Internet access through private NAT gateway is supported for DSW instances

Target customers:

  1. Customers who need to accelerate Internet access in scenarios such as model downloading.
  2. Customers who need to block Internet access for security reasons.

Features released: Internet access through private NAT gateway is supported for DSW instances.

Platform for AI - Encrypted NAS mounting supported in DSW and DLC

Data Science Workshop (DSW) and Deep Learning Containers (DLC) allow you to mount a NAS file system as an instance or a shared dataset.

Target customers: enterprise customers that have high requirements on data security.

Features released: DSW and DLC allow you to mount a NAS file system as an instance or a shared dataset.

Security Center - Threat Analysis Is Available in the Singapore Region

Data can be delivered to the China (Shanghai) or Singapore region based on compliance requirements.

Target customers: multinational enterprises or large corporations that want to use the threat analysis feature to switch between regions in the Security Center console. The log data of cloud services in the relevant regions can be delivered to the China (Shanghai) or Singapore region based on data compliance requirements.

Features released: The Singapore region is supported. Users can select the China or Outside China data management center to deliver log data. If China is selected, data is delivered to the China (Shanghai) region. If Outside China is selected, data is delivered to the Singapore region. Data that is generated by the threat analysis feature is also stored in the selected region.

Experience Enhancements

Elasticsearch - Plug-in Installation Cancellation Supported During Cluster Changes

You can cancel the installation of plug-ins during cluster changes in later versions such as V7.16, V8.5, and V8.9.

As for original control capabilities, you cannot perform operations during a cluster change. As for the current control capabilities, you can cancel the installation of a plug-in during a cluster change. This enables more flexible and convenient user experience and reduces O&M costs.

Content Moderation - Image Moderation 2.0 Supports the Multi-language Moderation of Text in Images

Image Moderation 2.0 supports multi-language moderation of text in images, which can better meet the moderation requirements of customers outside the Chinese mainland.

  1. Multi-language moderation of text in images is supported in the Singapore region ap-southeast-1.
  2. This feature can automatically identify the languages of text in images and detect violations.
  3. The following languages are supported: Chinese, English, French, German, Indonesian, Malay, Portuguese, Spanish, Thai, Vietnamese, Japanese, Arabic, Filipino, Hindi, Turkish, Russian, Italian, and Dutch.
0 1 0
Share on

Alibaba Cloud Community

875 posts | 198 followers

You may also like

Comments