At this Alibaba Cloud Summit 2020, Alibaba Cloud Native Group shared two topics. The following content is taken from the speech.
Prometheus & EDAS 3.0 Announcement
By Guoqiang Li.
Alibaba Cloud native team have made active contributions to the open source community, directly contributing to well-known open source software such as Apache dubbo, RocketMQ, Spring Cloud Alibaba, Dragonfly, Dragonwell, and actively participating in the construction of open source communities such as Kubernetes. At the same time, Alibaba Cloud also provides many cloud products that support open source software, such as container service Kubernetes, message queue RocketMQ, Alibaba service mesh, etc.
In the field of basic monitoring, the open source software prometheus is developing very rapidly, and it can comprehensively monitor the customer’s infrastructure, from virtual machines to components to Kubernetes. It can be well supported. With the release of prometheus service, until now, Alibaba Cloud has covered the field of application monitoring comprehensively, including metrics, tracing, and logging. From the host to the container, to the full stack support of the application, the language is also fully covered by open tracing.
The ARMS Prometheus service provides a fully compatible open source experience, supports PromQL, native service discovery and alerting.At the same time, Alibaba Cloud solve the pain points of native prometheus. It provides unlimited expansion of cloud storage, high availability and high stability, and very low load of Indicator collection consumption.
Pay-as-you-go, which is the payment model is cheaper than open source self-build. In addition, ARMS Prometheus provides dashboard and defaul alarms for out of box, which is more convenient to use. At present, ARMS Promethus and ACK are deeply integrated and are available in major regions around the world.
Guoqiang Li also introduced the latest version of EDAS. In EDAS3.0, Alibaba Cloud focused on strengthening the support for Kubernetes and the open source microservice framework Spring Cloud and Dubbo. When users deploy Java applications, they do not need to care whether the underlying layer is VM or Kubernetes. If the bottom layer is Kubernetes, EDAS will help users build an image and deploy it to Kubernetes. After that, the application’s full life cycle management, monitoring, logging, and microservice management capabilities are fully configured. EDAS3.0 also provides a series of critical capabilities in production.
In terms of microservices, EDAS supports all versions of spring cloud and dubbo in the past 5 years through a non-intrusive probe model. Users can enjoy the microservice governance capabilities provided by EDAS without deploying upgrades or modifying programs. EDAS also includes Alibaba’s safe production of three axes, observable, grayscale, and rollback, and all are built into the product design to ensure the user’s business stability. Based on the above capabilities, EDAS can support the operation of microservice loads on the cloud in a more cloud-native way, helping users build stable and efficient microservice operating environments.
Alibaba Cloud Native Solution
By Shuwei Yin.
Shuwei Yin (Lishan), the Senior Solution Architect for Alibaba Cloud-Native Services makes an Introduction to Alibaba Cloud-Native Solutions.
We all understand the importance of the digital transformation, especially in an era where a vast majority of business activity has moved online. To that end, businesses want to run and design their digital operations to deploy products faster, scale effortlessly, control costs, and sustain an innovative advantage. Cloud-Native is a major enabler of the digital transformation, and Alibaba Cloud-Native services can use their capabilities in elasticity, serverless, observation, and hybrid/multi cloud to fulfil these requirements.
Alibaba Cloud has a full and diverse portfolio of container services to support your cloud-native transformation. We have two primary container orchestration services, ACK, which includes dedicated and managed version of Kubernetes, and ASK, which is serverless (meaning you don’t manage any infrastructure and only pay for what you use). Built to support both public and private clouds, as well as edge computing, Alibaba’s container services are deeply integrated with high-performance compute, network, and storage products. On top, Alibaba Cloud provides end-to-end security, observability, and many more services to simplify, secure, and accelerate your cloud-native transformation.
Alibaba Cloud provides multiple different templates to choose from when creating your kubernetes cluster to meet your needs for a diverse pool of scenarios. For example, you can choose a heterogenous computing cluster template which uses NPU/GPU instances as worker nodes, to support your own machine learning and AI workloads.
There are two ways to run your kubernetes clusters on Alibaba Cloud. For long-running applications, you can use ACK to manage the pods running on basic ECS servers, but this will require capacity planning and manual maintenance. For temporary applications, like batch processing or CI/CD, you can use ASK to manage the pods running on ECI, where auto-scaling is done on demand and no node maintenance is required.
You can see that ACK supports a combination of different runtimes, RunC, sandboxed, as well as ECI. You can use a mixture of all three. Some pods can run in a normal RunC docker runtime, and some pods that require isolation and better security can run in a sandbox. You can use a combination of ECS nodes and serverless ECI nodes, whose pods will share the same attributes as a serverless Kubernetes cluster.
Heterogeneous computing is very important for AI and Machine Learning workloads. ACK is ideal for the job because AI algorithms must be able to scale to be optimally effective. Some deep learning algorithms and data sets require a large amount of compute. Kubernetes helps because it is all about scaling based on demand. It also provides a way to deploy AI-enabled workloads over multiple heterogenous resources across the software pipeline while abstracting away the management overhead.
Hybrid deployment has become a common choice for enterprises to migrate their workloads to the cloud. However, the adoption of hybrid cloud brings to mind a new challenge: There is a huge margin in terms of capabilities and security requirements between on-premises and cloud-based infrastructures. And so we arrive at the question: how can you manage both of them effectively at the same time? To address this issue, Alibaba Cloud’s Container Service for Kubernetes has provided the application-centric hybrid cloud 2.0 architecture.
You can use the “Register Cluster” feature for unified traffic and cluster management of on-premise and other cloud kubernetes clusters. This is such a great feature because you can easily manage your clusters from different cloud vendors, different runtime environments with unified governance, observability, scheduling, and deployment.
Alibaba Cloud container services are used by customers from more than seven thousand (7000) customers around the globe, from a variety of different industries. Many major brands like Siemens and Akulaku all run their most mission-critical applications on Alibaba Cloud container services.
This customer is Indonesia’s leading B2B marketplace lending platform for SMEs. This customer built the front-end and back-end platform through a monolithic architecture. As the modules and features of the application were increasing, the IT operation team faced a number of challenges, the biggest was the use of a monolithic architecture, which meant that any change request and upgrade had to involve editing the database, code, and front-end platform, which is quite a time-intensive process.
Since all individual components on the platform are highly coupled and dependent on each other, a single point of failure can bring down the entire system, leading to the outage of business operation.This customer deployed the application on ACK. Alibaba Cloud helped the customer redesign their monolith architecture to microservices, which involved breaking up the monolith into its own, single service components. Alibaba Cloud helped the customer re-architecture from monolith to microservices; shortened their time of App development and deployment; reduced IT and resources expenditures; and improved reliability and agility.
Adira Finance, established in 1990, is a leading Indonesia-based multi-finance company primarily engaged in provisioning consumer financing for either used or new motorcycles and cars. The company has close to 600 branches across Indonesia, over 21,000 employees, serving more than 3 million customers.
As their business grew, they needed to redesign their monolithic applications to microservice applications to increase business agility and cut IT costs. Alibaba Cloud stepped in and provided database, storage, log service, API gateway and other products to work in conjunction with ACK to support their new microservices architecture. Alibaba’s solution provided them with high availability with multi-zone redundancy to minimize business disruptions and allowed them to easily scale up or down, saving costs and boosting development and deployment efficiency.1_5_
Join us at the Alibaba Cloud Summit 2020 for more.