In a distributed architecture environment, the dependencies between services are increasingly complex, no one can say that the impact may manifest faults of the entire system, to build a highly available distributed system is facing a serious challenge.
July 17, Ali cloud application service availability AHAS official business, contains the schema-aware, flow control and error exercises downgrade the three independent functional modules, can rapidly improve application availability capabilities to solve problems in distributed high-availability architecture .
Live Review: Click Here
In a distributed environment, we often encounter a variety of components and complex dependencies, errors caused by enlarged partial failure, and instantly burst traffic due to congestion, the face of these challenges, we often difficult to start.
High availability service application AHAS full link from Alibaba pressure measurement technology, after years of double eleven exams Technology Group, providing flow control to downgrade the protection of strong stability and other means of protection. At the same time, provide an automatic topology-aware application, you can draw dependencies between components and applications rely on the infrastructure and provides applications to test drill scenario based real failure of high-availability capabilities, is the most chaotic works on the cloud good practice.
Flow control downgrade
This function is to control the assembly Sentinel (based Alibaba Group independently developed open source flow https://github.com/alibaba/sentinel ), Java applications and provides a gateway (Spring Cloud Gateway, Zuul) method of limiting the application level and downgrade , fit a variety of open-source framework, Agent and SDK support two access methods, improve application availability in the face of peak traffic, dependent on unstable services, with a variety of flow control, second-level real-time monitoring, low loss performance Features.
Fit a variety of open-source framework, Agent and SDK support two access methods to flow as the starting point, guarantee the stability of the dimension of flow control, fuse downgrade, system protection and other services, while providing a powerful second-level monitoring and historical monitoring queries.
Meanwhile, for the basis of SpringCloud, Dubbo or K8s micro-service applications, a key to quickly access high-availability capabilities, no modification of code without having a background operation and maintenance services.
Architecture perception
AHAS application high availability service provides automatic detection of intelligent recognition function, collection and analysis of standard operating systems and interfaces three parties, calling the relationship building process-level, feature-based database technology component recognition algorithm used by the process, and through the server, container, process three dimensional visualization display architecture.
Failure drill
By actively manufacturing fault, highly available application services AHAS can check the availability of the application system and its components behave under these faults, so the ability to verify the availability of applications, exposed to the risk of failure in advance, to avoid serious consequences.
According to reports, there are China Taiping, Amway, future markets and other business applications have high availability service AHAS deployed to the production environment to address application instability caused by unexpected traffic peak. It is worth noting, architecture-aware applications highly available and fault drills AHAS services offered are free of charge, while providing free access to each Region 5 current limiting demoted node every day, to help develop and operation and maintenance engineers a better experience key upgrade cloud on application availability.