Abstract:What kind of IT infrastructure can support high concurrent traffic that lasts a full day or more? How to avoid the pitfalls of 618? Listen to me carefully.
The special period has brought 2020 off an abnormal curve at the start of the New Year, with the social machine hit the pause button and fear spreading among everyone. After months of fighting and experience, the shadow of the virus is gradually subsiding and the 618 shopping festival is upon us. Every year on June 18, e-commerce companies compete with each other to make cakes and trains.
A very typical business demand in the e-commerce industry is the surge in traffic caused by promotional events. For example, in the 618 Campaign, the e-commerce system will encounter many challenges under the impact of high traffic volume. However, how to ensure that the IT system is highly stable and highly available in the order surge is a huge challenge for all 618 event organizers, which requires careful thinking and elaboration from technology, architecture and specific implementation measures. In daily business, e-commerce IT systems not only have high standards of system stability and security, but also need to deal with sudden traffic at low cost and provide flexible computing resources. These requirements drive the maturity of cloud technology.
With the development of technology and the transformation of enterprise digital transformation, it is found that the traditional server has been unable to meet the current business needs of enterprises. Traditional servers are prone to problems due to hardware reliability limitations. It also requires the purchase and deployment of various additional security measures, and cannot do multi-user to multi-server access control operations. If there is a problem, there is only one way to do it “manually”. Manual duplication of all deployments, manual backup, manual data recovery… . The last thing you do can make you doubt life.
So, what kind of IT infrastructure can support high concurrent traffic that lasts a full day or more? With more than 10 years of service experience in the e-commerce industry, Huawei Cloud offers an exclusive guide for e-commerce practitioners on how to avoid the pitfalls brought by 618. Wish you all the best with 618.
Under the peak traffic, users face unexpected service demands
The following problems may occur in the event of an explosion of user visits and orders during the 618 campaign
CPU spikes: Serialization and deserialization, high log output, and extensive reflection are the main reasons for CPU spikes. However, most core businesses have strict requirements on request response time, which puts forward strict requirements on the CPU performance of single instance.
Disk I/O bottleneck: Data buried and user behavior analysis are necessary to improve activity retention. However, a large amount of buried data leads to disk I/O surge, which affects service stability
Resource monitoring: How to identify capacity expansion requirements before reaching the system’s performance inflection point and load-bearing limit point? How to achieve mass monitoring of mass instances? How is monitoring data collected and stored?
In the face of the above problems, high-performance elastic cloud servers become necessary for enterprises to cope with traffic peaks
Huawei cloud in early 2020 to launch a new instance C6s aimed at high load scenarios, with stable performance and high cost performance, for the game, electricity, rendering and live video on server computing, storage, network capacity has a strong appeal to the scene to provide powerful basis for support, to help users easily cope with 618 surge in business activity.
General Computing Enhanced C6s, as a high-performance cloud server, is in the same line with C6 cloud server. It adopts a new generation of Intel processors, base band 2.6ghz, core band 3.4ghz, and memory frequency upgraded to 2922MHz, with excellent computing performance.
The C6s cloud server deeply optimizes the resource allocation algorithm on the virtualization layer to control the computing fluctuation within 5%. In addition, the C6s can be equipped with universal SSD cloud disks, improving the read and write performance by four times and reducing latency to 1ms. Even with high volumes of computation and extremely high concurrency, users can expect solid and robust computing performance.
In terms of network performance, the C6s is equipped with dual 25GE high-performance intelligent network interface cards (inics). The maximum Intranet bandwidth is up to 30 GB. The C6s supports high-performance network forwarding and the maximum packet sending and receiving capability is up to 8.5 million PPS. Lower network delay, higher internal and external bandwidth, and greater network throughput give C6s higher data transmission efficiency, enabling it to handle high load, multi-threading, large amount of data computing services with ease.
In terms of security, C6s integrates enterprise host security services by default. The account burst protection function provides fast and slow violence removal detection algorithms, and the IP blacklist of the whole network, which effectively prevents account violence removal attacks in a trinity. In addition, enterprise host security will precisely block the attack source IP for 24 hours after discovering the account burst attack and forbid it to launch another attack. At the same time, for host virus protection, the new provides known virus detection and high-risk automatic check and kill functions, greatly preventing host attacks from affecting user services.
Resource monitoring level, mainly is the host performance monitoring and network level, these are the basis of ensure the smooth running of business environment, with the cloud monitoring service, can view real-time running state resources and set the alarm rules, through in C6s Agent plug-in installation, to provide users with the server of the system level, active, fine granularity monitoring service. The data collection frequency is once a minute. In addition to CPU usage, it also supports memory usage (Linux). You can monitor active processes on a host by default, collecting information about the CPU and memory consumed by active processes, and the number of open files. Delay and error information should be reported proactively, and problematic services should be found in advance and related measures such as capacity expansion and downgrade should be taken
The super competitiveness of C6s cloud server is not only reflected in its excellent performance, but also reflected in its very competitive cost performance. Huawei Cloud constantly innovates technology to lower the threshold of high-performance products, enabling users to enjoy ultra-high performance products at the price of inclusive level. C6s is currently in the hot sale of Huawei Cloud 618 event, with discounts as low as 3.2 percent and 6600 gift packages, welcome to experience!
For small and medium-sized Internet enterprises and e-commerce enterprises, the construction of their own IT systems will be a large cost. Huawei cloud servers provide unified solutions for e-commerce enterprises and more smes to deploy applications on Huawei cloud platforms. Unified O&M and cloud platform security mechanisms ensure system stability and security, achieving an all-win pattern.
Click to follow, the first time to learn about Huawei cloud fresh technology ~