An exclusive interview with the winning team of OurBMC Open Source Competition is here!

The exciting OurBMC Open Source Competition has come to an end. After four months of actual combat, each participating team has also accumulated rich practical experience and participation experience. In this issue, the community specially invites the winning corporate teams of the OurBMC Open Source Competition to share "Enter the OurBMC Open Source Competition and jointly practice the open source spirit of openness, inclusiveness, co-creation and win-win", so that more people can see the charm of open source and the power of technology.

PART.01

· Entry background

As a board member of the OurBMC community, Inspur Computer actively responds to community activities and competes with BMC technology experts through the OurBMC open source competition to further enhance the BMC professional capabilities of team members.

· Core Program

In this award-winning work "Design of Server Fault Diagnosis and Prediction Platform Based on BMC Technology", the Inspur Computer team proposed a DTF (Dynamic Threshold Funnel) algorithm and a CPU high-temperature frequency reduction algorithm for "fault prediction". The DTF algorithm solves the problem of users frequently receiving CE (Correctable Error) alarms, and uses CE alarms to predict failures of fixed-position components and predict the health status of server components in advance. The CPU high-temperature reduction algorithm can assist in cooling the CPU. On the one hand, it relieves the heat dissipation pressure of the entire machine, and on the other hand, it also reduces a series of losses and negative impacts of the CPU due to high temperature.

· Participation process and experience

After receiving the competition questions, the team conducted an in-depth analysis of the competition questions and the intention of the question maker. At the same time, they did a lot of comparative analysis between the Feiteng platform and other platforms. They also consulted other internal platform teams, hardware, cooling and other teams within the company, and finally decided The CE fault and the direction of CPU high temperature and frequency reduction are the themes of the competition.

· I say to the community

Open source is an open, collaborative software development model that promotes technology sharing and innovation. The BMC technology stack is a comprehensive solution covering enterprise management, operation and maintenance, monitoring, etc., which can help enterprises achieve efficient, stable, and secure IT operations. The combination of open source and BMC technology stack will bring enterprises a more flexible, efficient and secure IT service experience and promote the digital transformation of enterprises. At the same time, open source also provides a broader space and more possibilities for the development of BMC technology stack.

PART.02

· Entry background

As a member unit of the OurBMC community, Kunlun Tech hopes to integrate the hard drive failure prediction function into the OurBMC platform through this competition, making the OurBMC system more complete and making its own contribution to the development of domestic BMC systems.

·  Core solution

This award-winning work "Design Description of Hard Drive Fault Prediction Based on OurBMC Platform" is mainly divided into fault model training and fault prediction. In order to improve the accuracy of prediction as much as possible, machine learning methods are used in the implementation of the solution to detect hard disk faults. predict. It conducts training and learning based on a large number of hard disk sample data, and finally forms a prediction model, and then uses this model and combines the current status information of the hard disk to make predictions. In this plan, the support vector machine algorithm was selected for training because this algorithm has the advantages of good classification effect and low resource consumption.

· Participation process and experience

After selecting the entry plan, the team conducted a careful and detailed analysis of the process framework for hard drive failure prediction, and also invited experts from each module to form a team. Many thanks to them for their hard work in this competition. This competition demonstrates the good cohesion and teamwork spirit of our team. In addition to winning the award, this is also an aspect that our team is proud of. At the same time, we would like to thank the company leaders for their strong support and recognition in all aspects of this competition.

· I say to the community

Everyone adds fuel to the flames. As a practitioner of BMC development, I hope to be deeply involved in the field of BMC and contribute my small contribution to the development of the OurBMC community. At the same time, we hope that our country’s independent and controllable BMC system will become stronger and stronger. In addition, I am very grateful to the Open Atom Foundation , OurBMC Community and Feiteng Information Technology Co., Ltd. for organizing this competition and providing a very valuable technology sharing and exchange platform for BMC technology enthusiasts.

PART.03

· Entry background

As a group of young people who love open source and are passionate about technology, we believe that competitions are not only a stage to showcase personal skills, but also an opportunity to learn, communicate and grow. Participating in competitions can not only improve one's technical level, but also recognize the gap with others, thereby building self-confidence and having the courage to face various challenges. This desire for challenges and the pursuit of growth drives us to keep moving forward and surpass ourselves.

· Core Program

This award-winning work "QMonitor Cluster Fault Diagnosis and Prediction Platform" is oriented to the Feiteng hardware ecosystem and is specially built for large-scale server cluster operation and maintenance management scenarios. Based on BMC technology, the platform extends server management from a single machine to a cluster, providing comprehensive monitoring, prediction and diagnosis functions. Use the concept of "phase space" to describe the overall activity and chaos of the cluster, use the quantum dynamics framework QDF and XGBoost algorithms to build a fault prediction model, achieve accurate fault prediction and diagnosis of the cluster, monitor hardware status, performance indicators and operating conditions in real time, and predict potential fault and alarm.

· Participation process and experience

Looking back on the entire competition process, the biggest difficulty is still in analyzing the competition questions. Since the entry requirements are relatively open, it has become a difficult problem to determine an entry plan that meets the requirements and is outstanding. Through the organizer's explanation of the competition questions and patient answering of questions in the communication group, our final plan gradually became clear. During the development and design process, most members of the team took care of their schoolwork and completed competition tasks at the same time. Although it was hard work, they were all very motivated to put what they had learned to use!

· I say to the community

The server BMC technology stack is an indispensable part of the modern data center. Related technical talents are in short supply, and there are few learning materials available on the Internet. For students, the learning threshold is high and there are few exposure opportunities. The OurBMC community continues to promote the development of BMC software and provides open source projects such as  OpenBMC and OurBMC, allowing users to manage servers more flexibly and customizedly. The development of this open source ecosystem not only reduces costs, but also promotes technological innovation and cooperation, bringing more choices and learning opportunities to enterprises and developers.

PART.04

· Entry background

As a pioneer in the field of localized independent controllable digital technology, we have been paying attention to community dynamics and activities since the establishment of the OurBMC community. It is hoped that through the open source competition organized by the community, the visibility of domestic BMC will be improved and more technical talents will be attracted to join the development of OurBMC.

· Core Program

This award-winning work "Remote Server Environment Collection and Early Warning Platform" mainly includes core functions and extended functions. The core function completes environmental monitoring, mainly based on the application layer dbus-sensors, entity-manager, phosphor-ipmi-ipmb, and redfish modules to monitor server temperature, voltage, current, etc., and implements log monitoring through log modules such as systemd-journald and rsyslog. . The extended function mainly involves four aspects: ① Use  the UOL  solution to realize the interaction and storage of server debug data and BMC. ② Use dual firmware backup scheme to realize abnormal switching of firmware startup. ③ CPLD resource expansion and CPLD WDT watchdog reset server and BMC. ④ PCle shares space with BMC memory Bar to implement server failure analysis or AI functions.

· Participation process and experience

First of all, I would like to thank the organizers of this competition. Their hard work and careful preparation provided a solid guarantee for the smooth progress of the competition. Especially in the preparation stage of the competition, the staff held a training session on answering questions, which was very important for us to understand and master the competition questions. In the competition, I think the most challenging thing is how we can accurately and concisely extract the highlights of the work within a limited time. This is not only a test of our professional abilities, but also a challenge to our thinking agility and creativity. At the same time, we need to improve efficiency and maintain our original intention to better balance the relationship between competition and work.

· I say to the community

Open source is the best platform and model for human technological progress. It plays a key role in identifying and filling deficiencies in software and hardware systems and improving the security and reliability of the system. The development and construction of OurBMC community has a vital impact on our domestic BMC technology, ecology and regulations, and plays an important role in catching up with foreign technology. We hope that more people can join the OurBMC community to build an innovative platform and promote the prosperity and development of the industry ecosystem.

PART.05

· Entry background

The recent BMC project has encountered technical difficulties. I hope that through the BMC Open Source Competition, I can get the opportunity to practice BMC adaptation, communicate and learn with other developers who are interested in BMC, solve problems encountered at work, and promote the development of BMC technology.

· Core Program

This award-winning work "BMC-based hard disk fault diagnosis and prediction technology design", the workflow is as follows: first compile the OurBMC system and transplant it to the Feitengpai file system; secondly, develop the mctp driver based on i2c and based on The mctp driver is transplanted to the nvme cli command line tool to obtain the information of the nvme hard disk; thirdly, it is judged whether there is a fault based on the information of the nvme hard disk, and the wear value information obtained is used to determine how long the hard disk can be used. This solution can improve the OurBMC system's hard drive fault diagnosis capabilities and avoid serious system problems caused by hard drive failures.

· Participation process and experience

During the competition, we encountered a series of challenges including time constraints and technical difficulties. Faced with the dual pressure of finishing the work project and submitting the competition work, the team could only devote a lot of evenings and weekends to studying and completing the work. For BMC technology and embedded system development and transplantation that I have little exposure to, I can only learn and use them now. It is also these challenges that have trained our team members’ emergency response capabilities, speed of learning new knowledge, and problem-solving abilities. Not only did I gain practical experience and improve my skills through this participation, I also felt the strength and warmth of the community.

· I say to the community

With the continuous development of BMC technology, more and more devices need to be connected to the BMC technology stack for management. In addition to complete machine manufacturers, IBV and chip manufacturers, the community can demonstrate the advantages and value of BMC technology by holding industry exchange meetings, technology matchmaking meetings and other activities, and attract more equipment manufacturers to join the community. At the same time, we hope that the community can provide more abundant tutorials, documents and cases so that more developers can learn and learn from it.

The first OurBMC Open Source Competition has come to a successful conclusion, but open source continues and innovation continues. We look forward to more developers becoming brave explorers of the open source world. OurBMC community sincerely invites colleagues in the industry to actively participate in community construction and exchanges, contribute to the community and BMC ecology, and achieve more outstanding achievements together.

A programmer born in the 1990s developed a video porting software and made over 7 million in less than a year. The ending was very punishing! High school students create their own open source programming language as a coming-of-age ceremony - sharp comments from netizens: Relying on RustDesk due to rampant fraud, domestic service Taobao (taobao.com) suspended domestic services and restarted web version optimization work Java 17 is the most commonly used Java LTS version Windows 10 market share Reaching 70%, Windows 11 continues to decline Open Source Daily | Google supports Hongmeng to take over; open source Rabbit R1; Android phones supported by Docker; Microsoft's anxiety and ambition; Haier Electric shuts down the open platform Apple releases M4 chip Google deletes Android universal kernel (ACK ) Support for RISC-V architecture Yunfeng resigned from Alibaba and plans to produce independent games on the Windows platform in the future
{{o.name}}
{{m.name}}

Guess you like

Origin my.oschina.net/ourbmc/blog/11094425