Competitive PK of millisecond level: 5g speed needed by telecommunication industry
voltdb 2021-06-04 09:35:51

VoltDB It's a high-speed decision engine , Proven to power real-time applications , These real-time applications have to respond in milliseconds , To increase revenue or prevent revenue from decreasing .

In the telecommunications industry , This kind of core competence is absolutely necessary .5G The introduction of Internet of things (IoT) Combined with the explosive growth of , Driving the amount of data 、 A massive increase in speed and variety , because 5G Driven service level agreements (SLA) The power gap between requirements and legacy system weaknesses , To reach scale , Traditional solutions alone can't do it 5G The latency and transaction consistency level .

VoltDB The unique functions of the platform make it better than other platforms to handle the complex high-speed data needs of telecom companies , Digitize your business . This is particularly effective for the following application use cases :

  • Digital business support system (BSS) Unable to meet 5G New delay requirements for , New modular deployment methods are needed .
  • Billing Intermediary system , Due to the inflow of all IoT The data of , So there are huge new opportunities for development .
  • Customer management ( Also known as customer value management and customer experience management ) In fact, it has developed into an industry , Its highly personalized quotation has become its standard .
  • Income guarantee , In the face of growing cyber threats and more complex fraudsters, it's becoming more and more difficult .

VoltDB It aims to increase the big data investment of telecom enterprises in advance ( for example NoSQL,Hadoop And traditional data warehouse ) And news investment ( for example Kafka), So that the enterprise can trigger the correct decision in the key events on a large scale for the complex stream data . It only takes less than 10 Time in milliseconds , And without affecting the accuracy of the data , Increase revenue or prevent loss of revenue in real time .

millisecond -VoltDB The advantages of

The difference between the world's top sprinters and the endless Olympians is a few milliseconds . Again , The data world has become a millisecond game , Losing a fraction of a second means losing revenue and losing to competitors .

VoltDB I know that better than anyone else , That's why we build our platform around the combination of speed and accuracy : We didn't compromise on either .
VoltDB Designed to support real-time applications , These applications have to respond within milliseconds of the event , To increase income or prevent loss of income .

Please note that , We're talking about “ Incident ”, instead of “ After the event ”(“ event ” It's the act of taking steps to trigger the transmission of data over fiber optic cables , For example, credit card purchases , Texting or being told to do something through a phone app IoT equipment ).

We will “ Incident ” Defined as after the event 10 In milliseconds , And before sending the data involved to the data Lake .

1. Why is the concept of unit millisecond so important ?
If the data has reached the data Lake , It's going to lose the opportunity to cash in , And lose most of the value , And then lost the monetization (BSS) The ability of , And the loss of using it to attract customers' attention ( Customer management ) The opportunity of .

2. Why is it 7 To 10 Missed these opportunities in milliseconds , Data has lost value ?
Every task from loading a web page to making an online purchase contains many subtasks , Each subtask has its own delay , If your data platform can't meet the requirements , These subtasks end up taking too much time , Cause shoppers to cancel orders or cyber criminals to sneak in .
This is it. VoltDB A place of great splendor . If the whole process needs to be in 250 Complete in milliseconds , that “ Decision making ” part - When your database through the use of machine learning and AI To determine whether a piece of data is abnormal , So it's worth taking action ( adopt , Or don't let it go through , Or use it in time to make a profit )- Need to be in 10 In milliseconds , So that all the subtasks around it happen in time .
When you think about 5G and IoT when , It's not enough to do this for abnormal data 10 Millisecond decisions become more valuable , This will make communication between devices faster . Because the speed of data is limited by the speed of light , The speed of light travels through a medium that travels at the speed of light ( optical cable ) Further reduction , So speed up processing and make the most of 5G The only way to do this is to do as many data events as possible ( That is to say “ edge ”), From ingestion to proper operation , To get the data as quickly as possible .

Advantages of features - About VoltDB

VoltDB Provides a rare combination of features , Make our data platform particularly suitable for 21 Century Telecom and all kinds of software suppliers providing services for it .
Today's business environment requires some hard core capabilities to address the scale and latency challenges of specific applications , Without affecting accuracy or reliability , and VoltDB Each of the following functions can be provided :
First, the application level features and benefits :

Watch it 1. Ultra reliable low latency , But with 5G SLA bring into correspondence with

5G And IoT And machine to machine communication , Make traditional technology almost impossible to survive . In order to satisfy the 5G Of SLA, In especial BSS And customer management requirements , Your application must be able to 200 From receiving to making decisions in milliseconds , This means that the decision-making part needs to be in 10 Complete in milliseconds , At the same time, it does not affect the accuracy of the data .

Watch it 2. Stream integration , To make decisions in an event based on multiple information flows

Today's data comes from highly dispersed sources , fast .
Stream processing technology enables you to process in real time 、 Storage 、 Analyze the data , most important of all , You can manipulate the data stream . Traditional batch data processing method , The request is in the process of 、 Collect data in batch before storing or analyzing it , And stream data flows into , So that you can process data in real time , Instead of waiting for the data to arrive in batch .

Watch it 3. Data processing that follows strict standards to keep transactions consistent , And make complex decisions

Adherence to strict standards is the key to consistency .
If 5G Make it all faster , So while ensuring consistency without sacrificing other metrics such as speed or data platform performance ,ACID Become more important , It's also becoming more difficult .
Most are based on NoSQL Our data platform will not be able to meet one or more stringent requirements , So it can't satisfy many of today's ( Even if not most ) Data consistency requirements for enterprise applications .

Watch it 4. Intelligent real-time aggregation , Can be measured , Monitoring and detection are important KPI Deviation in

Any powerful 5G The core of the application is the ability to use history and recent records to make complex decisions , So as to have a subtle understanding of the event , And according to your unique business needs and KPI Act on it in the best way .

Most data platforms ( Especially those who use older versions or NoSQL Data platform of database technology ) Focus on historical data or the latest data , Instead of summarizing it and getting an overall view .

Watch it 5. Horizontal scalability , To support the growing amount of data and all kinds of data , Without delay

Due to the overhead involved and the data packets, additional data center transmission must be carried out , So just vertical expansion ( That is, by adding CPU kernel ) It can quickly cause your application to delay and run longer than normal .
Memory data platform ( Such as VoltDB) Optimized for horizontal and vertical expansion , Because they're not bound to disk , And you can use partitions to scale horizontally , You can also use multi-core support to scale vertically .
besides , There are also some operation level features and benefits :

1. Cloud native will build competitive advantage through agility and Automation

Cloud native architecture makes full use of the distributed nature of public cloud 、 Extensibility 、 Flexible features , To maximize business value and maintain customer satisfaction .
Going to the cloud business means enabling practices such as continuous integration and deployment, and the ability to quickly scale or reduce resources to real scale , Reduce many layers of infrastructure ( The Internet 、 The server 、 Operating system, etc ) To accelerate the development and deployment of new services . Real time resource optimization , In response to peak traffic and one-time events .

2. High availability ensures business continuity in the face of hardware failure

stay 5G Time , Customer management 、BSS And revenue must be required without service breakdown . To keep your customers happy and loyal , Before fraudsters have a chance to wreak havoc in your network , You need to avoid them .
With high availability , You can reduce the negative impact of downtime , Realize the automatic recovery of system failure , And then into a better return on investment , And ultimately achieve a more robust bottom line .

3. Take the initiative - Proactive replication across data centers , To provide the flexibility of geographical distribution , To support the 5G Quality of service

While many data platform and database technology companies offer some form of cross data center replication , But very few people provide real live cross data center replication .
This means using two databases , Both databases can be changed in real time , And both databases can propagate changes to each other , So the problem of conflict resolution is avoided , Because these problems significantly affect application performance and ultimately your return on investment .

4. Disk persistence ensures that you never lose data

Disk persistence means that at the end of the data creation process , The data will remain .
Disk persistence ensures data storage when the system you are using is powered off , This means that you can get the most value from the data , Without losing any data .

5. By reducing the number of layers , Simplify the infrastructure , Simplify the stack to reduce costs and improve data processing performance

Each layer in the stack adds latency . Adding more layers can solve all kinds of data ingestion and analysis problems , And applications can no longer spend time making mission critical decisions .
until 5G and IoT Appearance , Changed the status quo . The fewer layers , The less delay you have to introduce the stack .VoltDB Allows you to simplify the stack , To manage 200 In milliseconds or less “ From instant access to operation ”.
Although many database technology companies offer some of the above “THE” function , But no other data platform or database technology company provides all the above functions .

VoltDB Strong support ACID And real-time intelligent decision-making applications , To make the world connected . There is no other database product like VoltDB such , Low latency can be required at the same time 、 On a large scale 、 High concurrency and accuracy of the combination of application refueling .

About VoltDB
VoltDB Strong support ACID And real-time intelligent decision-making applications , To make the world connected . There is no other database product like VoltDB such , Low latency can be required at the same time 、 On a large scale 、 High concurrency and accuracy of the combination of application refueling .
VoltDB from 2014 Turing prize winner in Mike Stonebraker Doctor created , He redesigned the relational database , In response to today's growing challenges of real-time operation and machine learning .Stonebraker Doctor has been studying database technology for a long time 40 many years , In fast data , Streaming data and in memory databases bring a lot of innovative ideas .
stay VoltDB In the process of R & D , He realized the full potential of using memory transaction database technology to mine streaming data , It can not only meet the delay and concurrency requirements of data processing , It also provides real-time analysis and decision-making .VoltDB It's a trusted name in the industry , At NOKIA 、 The financial times 、 Mitsubishi Electric 、HPE、 Barclay 、 Huawei and other leading organizations cooperate with each other in real situations .

VoltDB China :https://www.voltdb-china.cn/

Please bring the original link to reprint ,thank
Similar articles

2021-08-09

2021-08-09