Abstract: Cloud computing brings the change of IT goes without saying that as a new type of IT delivery mode, and to save IT costs of enterprises, speed up IT combined with the enterprise business efficiency, promote innovation ability, strengthening the management level and enhance the reliability of the system itself and so on to provide great support, is an important way of promoting the enterprise to realize the new development, IT has become the mainstream of the global IT industry.
Click here to view the original text:click.aliyun.com/m/42757/
Cloud computing brings the change of IT goes without saying that as a new type of IT delivery mode, and to save IT costs of enterprises, speed up IT combined with the enterprise business efficiency, promote innovation ability, strengthening the management level and enhance the reliability of the system itself and so on to provide great support, is an important way of promoting the enterprise to realize the new development, IT has become the mainstream of the global IT industry.
Therefore, CSDN reporter recently interviewed the first domestic provide maps API gould’s senior operational experts was the Internet service provider, share the gold from the perspective of ops on cloud behind the story and the benefits of the cloud, such as the change of the architecture, the core value of operational finally achieved application service availability up to 99.99%, compared with the improved 5 times, And the number of operations staff dropped from 45 to 16. He also said that enterprises should timely seize the new opportunities brought by cloud computing technology, in view of their own shortcomings, the implementation of the corresponding technology and business reform, embrace cloud computing, give full play to the role of cloud computing technology in enterprise management, so as to enhance the competitiveness of enterprises.
Autonavi’s ROAD of IT informatization
Autonavi was founded in 2002. Over the past ten years of development, Autonavi products have been applied to the entire Internet from simple vehicle navigation products. At present, the products or services provided mainly include basic services: positioning, search, route planning, navigation, walking and public transportation. O2O: hotels, food, scenic spots, etc. Internet of vehicles: AMAP AUTO version, driverless; Autonavi Big Data: real-time traffic conditions, congestion avoidance, etc.
However, today’s powerful products or services have also experienced a long IT information growth process. From the perspective of operation and maintenance, there are the following stages: chaotic stage without operation and maintenance team. From 2002 to 2007, I was in the primary IT stage. The IT department was responsible for the data production server, while the R&D staff was responsible for the IDC server, which was mainly used for the production of map data. Only a small part of the online map was available. Start to establish operation and maintenance team, professional operation. From 2007 to 2014, IT informatization was mainly divided into IT and operation and maintenance. IT was responsible for the product business of office network, including office computers, printers and office process software. The overall operation and maintenance is in charge of IDC room, server and business of the company. The overall operation and maintenance is still in the era of human flesh, and the efficiency of operation and maintenance service is low. From July 2014 to present, Autonavi on cloud has realized automatic operation and maintenance. At this stage, Autonavi began to go cloud, and operation and maintenance gradually stepped into the era of automation. The fault handling ability and performance of operation and maintenance personnel were further improved.
Autonavi’s data acquisition, processing and integration process
From the initial data acquisition, synthesis and geographical changes to today’s vehicle positioning system, Autonavi’s data processing and integration are mainly divided into the following stages:
Primary stage: The acquisition vehicle installed cameras and GPS, real scene acquisition, mailing video disc and other original ways; In addition, the internal operator performs data analysis and production by watching video and tracking, thus realizing data update once every three months. In this stage, it is all manual mode and the Internet is low;
System stage: the acquisition vehicle starts to use the acquisition system software, and then transmits back through the network in the evening, breaking away from the form of optical disk operation; After that, through the perfect editing tools, inspection tools, to achieve high quality control and fast assembly line speed, so as to enter the data processing stage;
Platform stage: The online high-precision acquisition vehicle collects panoramic and decimeter-level high-precision track, and carries out social collection in the form of crowdsourcing; Internal work platform, data flow task flow platform, and the whole process is monitored, and finally into the assembly line production stage;
Automation stage: In the future, the collection will be fully crowdsourced and automatic transmission will be realized. The automation level of the internal industry will be highly improved. The automatic production level will be improved through image recognition, big data analysis, automatic data fusion and other technologies, which is also the direction of Autonavi’s current efforts.
At present, Autonavi mainly has two App products, map and navigation, as well as PC map. At the same time, it also opens API services to facilitate mobile developers to better provide users with high-quality location-based services.
The pain of operations
In the cloud, before Scott has experienced a lot of pain in operational aspect, such as the self-built room of ops, Scott before seven small rooms scattered distribution in various places across the country, but with the rapid growth of the business to build a small room brings a lot of problems, including IDC power, air conditioning, the network is easy to malfunction, also makes the time of the shipment vader to more than 40 people; At the same time, it also faces a lot of operation and maintenance requirements, Because o&M needs to be responsible for IDC, power, network, server, OS, database (MySQL, Redis, Oracle, Memcache, PG, MongoDB), various application software (Nginx, Apache), open source tools (Jenkins, Puppet), A large number of demands make the experience of operation personnel scattered, resulting in a great deal of pressure. In addition, sometimes due to the heavy traffic caused by the user surge, coupled with the already tight bandwidth and other issues, even caused major events/festivals downtime.
Based on ali Cloud best operation and maintenance practices
Zeng Wei said that autonavi’s pain and burden in operation and maintenance have been fundamentally changed after The launch of Aliyun. First, aliyun has realized the operation and maintenance of unified infrastructure, so that operation and maintenance personnel no longer operate and maintain infrastructure. At the same time, Ali Cloud provides a variety of cloud services including load balancing, elastic computing ECS, cloud database RDS, object storage OSS, cloud database Redis version, big data computing service MaxCompute (original ODPS), etc. Operation and maintenance personnel only need to make good use of these services and do not need to maintain specific components. In addition, Autonavi often has large-scale activities, and user visits will surge at a certain stage. At this time, the flexible and scalable services based on Ali Cloud will be needed to achieve rapid expansion of infrastructure. Finally, the o&M staff only needs to focus on the core values of the business, such as performance, architecture, and failures.
However, at the beginning of Autonavi’s cloud service, operation and maintenance personnel familiar with traditional data centers may not be familiar with various cloud service components, and real-time log transmission delay may lead to the delay of real-time data analysis during the online operation of AUtonavi IDC and Ali Cloud. Through the perfect technology, timely response and training support provided by Ali Cloud, problems were quickly solved, and finally operation and maintenance personnel could skillfully use cloud services. For example, in the process of transferring MySQL database from Autonavi to RDS of Ali Cloud, we encountered problems of data transfer to cloud — data format, writing mode, export/import mode. Finally, the POWERFUL RDS of Ali Cloud was used to build DTS system directly through the transfer zone, realizing seamless switching and data consistency of cross-network database migration. In addition, MAO Yiming, technical expert of Alibaba Cloud, also introduced that Alibaba has also used its own Data Replication Center (DRC) product, which supports real-time synchronization of heterogeneous databases, Data record change subscription service, and provides solutions for cross-domain real-time synchronization, real-time incremental distribution, remote hypermetro, distributed database and other scenarios.
After moving to the cloud, amap’s app service availability rose to 99.99%, a fivefold improvement over the previous seven small rooms. In addition, from the previous system collapse of each large-scale activity, to the National Day travel peak operation and maintenance staff calm, the system is like a rock, fully reflects the stability of Ali cloud computing. Furthermore, autonavi had to purchase servers, put them on shelves, install them, expand human flesh and release them at least 2 months in advance for the launch and expansion of new applications. After the launch of Aliyun, the flexible expansion and reduction of capacity can be basically completed within 20 minutes, and the application of resources can be completed within 10 to 20 minutes for the launch of new applications.
Now, all of Autonavi’s core business systems have been put on the cloud, and Ali Cloud provides computing services. In terms of specific operations people, they don’t have to operate and maintain infrastructure — IDCs, networks, servers, etc. All kinds of database, cache, load balancing, publishing tools do not have to maintain, use ali Cloud services; You only need to apply for cloud services in different regions for remote Dr. You do not need to build equipment rooms. In terms of overall security, Ali Cloud provides many services, such as anti-attack (DDos), traffic cleaning, anti-virus, anti-fraud, code scanning, etc. Yiming MAO summed up the services provided by Aliyun for Autonavi in the following aspects: First, IaaS (Infrastructure as-A-Service), including multi-site computer rooms, network, inspection, etc. The second is PaaS (platform-as-a-Service). Autonavi uses cloud products including elastic computing ECS, cloud database RDS, object storage OSS, load balancing, Content delivery Network (CDN), and Big data computing Service MaxCompute(original ODPS). Finally, there is software-as-a-Service (SaaS) : services provided by Autonavi’s own systems (see chart).
Overall, for goldwin d the biggest change is the change of operational mode, from the original IDC to basic services such as network to the database to their maintenance, change to use ali cloud of cloud services, operations staff can focus on core competencies, such as architecture, fault, performance, a lot of improved its employees a sense of accomplishment, At the same time, the number of operations staff has been reduced from 45 to 16. In other words, the comprehensive cloud is the transformation and reform of the core value of operation and maintenance.
Finally, Zeng Wei also introduced that, in addition to conventional navigation services, Autonavi uses the powerful computing processing capacity of Ali Cloud’s “data plus” platform, such as Ali Cloud’s big data computing service MaxCompute(original ODPS), StreamCompute and other services. Autonavi can make route selection, congestion big data, congestion avoidance, urban hot spot distribution, scenic spot recommendation, and business area heat. It mainly conducts real-time data mining based on users’ positioning data, so as to develop a prediction model and obtain real-time results useful to users.
If the road condition is predicted at 4pm on a certain day, it is to get the historical data first, train and make an offline historical regression and prediction model every day, and then adjust and optimize the online model, so as to achieve the optimal prediction. In the face of future intelligent transportation and Internet of Vehicles, Decky launched AMAP AUTO version in Internet of Vehicles, which solves the problems of slow data update, poor user experience, lack of real-time road conditions and insufficient intelligence of traditional in-vehicle systems. In the long term, driverless driving is also a very important direction, but behind the Internet of vehicles, It is Also Autonavi that has taken full advantage of the big data capabilities of cloud computing.
Scan for more information:
