ZEGOCLOUD hits 2 billion minutes in the daily average interactions through its “X+audio/video” aPaaS access to all industries.
In response to remote service needs that exploded in various industries during the pandemic, audio/video cloud service provider ZEGOCLOUD instantly unveiled two SaaS platforms. These include TalkLine video conference and Xiaoyibang online help in preparation for entrance examination for art majors and many remote aPaaS solutions.
ZEGOCLOUD technology
Taking advantage of the strengths of download-and-use and quick development, ZEGOCLOUD rapidly broadened the applied scenarios of audio/video technology. From Live broadcasts, social networking, and an-entertainment to education, healthcare, and finance, public service areas.
As the collaborative online remote service apps were exploding, ZEGOCLOUD as their underlying technical service provider kept building up its platform database as well. The vendor’s total platform stream increased nearly 4 times and the daily average duration of audio/video interaction rose to 2 billion minutes earlier in 2020.
Real-time audio/video products and services cover real-time voice and video calls, mic-to-mic interaction, live broadcasts, and much more.
Partnerships
ZEGOCLOUD’s partners cover a mass of top pan-entertainment and online education platforms. Among them Inke, Ximalaya, Kugou, Lizhi, TAL, Zuoyebang, 17zuoye, or Zhangmen 1V1. The two industries were exactly the fastest-growing industries regarding data traffic earlier in 2020. The user count of many platforms was mounting by multiples, among which, some top national online education platforms.
The sharp increase in data traffic indicated a mass of users to platforms but caused greater pressure on platform stability. Thus leading to platform collapse and service unavailability. In those “stay-at-home” days, we all experienced the crash of many Apps.
How did ZEGOCLOUD achieve these numbers?
How did ZEGOCLOUD assure service stability given a sharp increase in data traffic, 2-billion-min-long interaction per day on average, and concurrence of thousands of millions of data requests?
Ten-million-level high concurrency architecture + backend service mechanism assuring service stabilityTen-million-level high concurrency requires an underlying architecture available for ten-million-level concurrency. ZEGOCLOUD’s founding members were Tencent’s former employees, most of whom got involved in fulfilling a leapfrog increase in the QQ product user population from millions of persons to billions of persons. Therefore, they have foresight in high concurrency design on underlying platform architecture. The required concurrency attained the ten-million-level in underlying architecture design as early as 5 years ago at the inception of ZEGOCLOUD.
Constant improvements
In subsequent years, ZEGOCLOUD relentlessly improved the technology and optimized the single point load capacity of servers, intelligent dispatching algorithm of the trunking system, and multi-cloud and -net virtual network building. As a result, ZEGOCLOUD has successfully put into practice flexibility. Now its users can experience seamless and smooth capacity expansion/reduction in a damage-free and unwitting way. In June 2019, ZEGOCLOUD’s daily average audio/video interaction totaled 600 million minutes in backend statistics.
Apart from the ten-million-level concurrency infrastructure, such backend service mechanisms as a warning, data traffic limiting, and emergency capacity expansion sustain service smoothness and stability upon waves of traffic peaks.
In this respect, Jack heading ZEGOCLOUD’s backend media service team elaborated on the operation of ZEGOCLOUD’s package of backend service mechanisms: for every enterprise customer, first of all, ZEGOCLOUD provides real-time data traffic monitoring and warning services. Backend service technicians monitor the current capacity level. When the customer’s server load exceeds a certain threshold, the automatic trigger of warning reminds those concerned of capacity expansion.
Customer service
Secondly, in the event of an instant sudden sharp increase, ZEGOCLOUD’s backend service staff will assure the availability of customer service through the data traffic limiting mechanism. What needs to be noted is that ZEGOCLOUD’s data traffic limiting mechanism is premised by no impact on the user experience. Take one of ZEGOCLOUD’s customers, a certain education platform, for example. Unexpectedly, the customer found a large jump in the platform’s user count. Then ZEGOCLOUD’s data traffic limiting mechanism was triggered. After data traffic limiting was enabled, the platform’s existing teachers and students sustained no impact on service experience; and for new users, ZEGOCLOUD sacrificed part of the user experience in exchange for service availability by lowering bitrate and resolution in stream publishing/playing. Subsequently, ZEGOCLOUD immediately enabled the emergency capacity expansion mechanism. Ten minutes later, capacity expansion was completed and new-coming teachers and students experienced a recovery in the bitrate and resolution.
Emergency capacity expansion
ZEGOCLOUD’s emergency capacity expansion mechanism is available for increasing the capacity of hundreds of servers in 10 minutes. And such server resources can support an online education platform to add programs involving 100,000 teachers offering courses or nearly 1 million attending students. Well prepared in the architecture and backend service mechanisms, ZEGOCLOUD also renders one-to-one technical support to help customers avoid beforehand and timely address problems that may arise out of data traffic increase.
Pre-services + one-to-one support helping all ZEGOCLOUD users gain the best experience”Technical support is a meticulous and thorough job and a sharp increase in the platform’s user count is a daunting challenge to it. We all are 24 hours on call to support increasing user count and expanding capacity for customers throughout the entire process,” quoted Tina, ZEGOCLOUD’s technical support director. Different from the normal image of a technical supporter that “solves problems that come up”, ZEGOCLOUD advocates mapping customer success orientation to technical support work or providing pre-services to help customers evade from possible risks in advance.
ZEGOCLOUD challenges
Behind 2 billion minutes of interaction per day is a hefty number of customers well recognizing ZEGOCLOUD and its services. But it did encounter some intractable problems. As online users were mounting at the fastest pace, server resources constrained the entire network. On one hand, customers were badly thirsty for capacity expansion to support service; on the other hand, vendors’ resources did not suffice. After taking possession of all available resources from the existing vendors, ZEGOCLOUD purchased from other cloud service providers one after another.”Our architecture is already in favor of multi-cloud e-commerce solutions, compatible with and supportive of physical machines and cloud hosts made by different manufacturers.
Therefore, our server resources go beyond single vendors and we can call as many resources as possible in no time.
Conclusion
ZEGOCLOUD highlights values of altruism and common growth and aspires for giving full support to customer success and growing with customers. In line with the principle of “perfecting technology and not wasting every penny spent by customers”, ZEGOCLOUD optimizes each customer’s cloud service cost fundamentally through relentless optimization of server load capacity.
In addition, for different customers at different periods, ZEGOCLOUD introduces different preferential measures. During the outbreak of the pandemic, users had a mass of requests for telecommuting and office automation (OA). In response, ZEGOCLOUD’s video conference platform TalkLine endowed every newly registered user with 5000 minutes of service free of charge. So far, approximately 1 billion minutes of service have been presented for free in accumulation.
When the wave arrives, ZEGOCLOUD gives the greatest support to customers; after the wave ebbs, ZEGOCLOUD makes a thorough arrangement for customers.
Talk to Expert
Learn more about our solutions and get your question answered.
