TechCasts

Read the Transcript

Amani: Hello everyone and welcome back to SYSGO Techcast. So last time we explored edge-to-hand cloud computing and we talked about how PikeOS helps make systems safer and more secure. And today we are diving into another critical piece of the puzzle, which is redundancy. It is the behind-the-scenes hero that keeps systems running smoothly even when things don't go as planned. So speaking to you is Amani and today I'm joined again by Dr. George Violettas. So let's start with redundancy. Can you please judge explain how it is integrated into cloud computing systems and why redundancy is so crucial today?

George: Yes, indeed. Transparency is one of the major topics, not only in cloud computing, but in all aspects of computing. And it has multiple different faces and implementations. First of all, the number one safe redundancy policy mechanism is the geographic distribution. Why? Because the things that we cannot even touch, we don't know when and how they're going to strike, are what we call force majeure or acts of God. And they include all kinds of natural disasters like flood, fires, earthquakes. And because those things are unpredictable, the only way to make sure that we face them is by distributing the storage in geographically distant locations. More than, if I remember well, it's more than 80 kilometers. So whatever natural disaster happens into one place, the other place is kind of safe. Of course, inside each of those data centers, we're talking about redundant hardware. So we have multiple hardware copies, ideally of different architectures. So one RISC-V, one arm, one x86. We'll compose a cluster. They will all do the same work. And we'll have some kind of voting mechanism, some kind of gateway in order to make sure that the data in and out are the correct ones. In order for those clusters to communicate, we're talking about network redundancy. So we need multiple even cables, even hardware connections, even multiple cables, but also network devices to reach the redundancy-required level. Switches, hubs, routers, etc. And of course, above all, we're talking about regular backups, usually automated. So whatever data we have, we have to back them up because you never know which storage device will be corrupted, because you need to make sure that you have the data integrity needed or even a cyber attack. For example, ransomware can be easily solved with offline backup, yeah? Because ransomware does not steal the data, it just locks the data. So if you have an earlier backup somewhere stored, you can just restore your data. That's it. Problem solved.

Amani: So now that we understood how redundancy is structured, let's address the topic of business continuity. Could you maybe dive deeper into how these strategies enhance both safety and security of cloud environments and what are the practical impacts of redundancy on system performance and business continuity?

George: Yeah, I mentioned some of those. Let's go deeply into each one of them. Geographic redundancy: As I said before, we're talking about different locations, physical locations, more than 80 kilometers away, which covers us from all natural disasters. Or even, let's say, regional power outages or whatever else could happen to one location. The other one will not be affected. 
Hardware network redundancy, which means that we have multiple servers, multiple devices and network paths, network cabling within the same data center. So there is no "we have to eliminate a single point of failure". If one of those, if whatever of those hardware components fails, we have to make sure that the other one, there is another one, there's a copy, a twin, which will take over. So the availability of the system will not be touched. 
Failover mechanisms: Where we have to monitor for failures and automatically threats, operations. Backup systems can be deployed even automatically. So the downtime has to be minimized. There are several barriers of which downtime is acceptable, depending on the safety and the security standards that the system we are running needs. As I mentioned before, backups and disaster recovery plans, each modern new business needs to have a disaster recovery plan. So not only regularly backup all the sensitive data, but make sure that they have a plan how to restore those data. If the building was damaged, if the computer room was damaged or even if some hardware was damaged, they have to have a plan on where and how to recover the regular backups. All those have a direct impact on the system performance and the business continuity of the company, including reliability, redundancy, enhances the system reliability.

And we need to make sure that security is kept up to certain standards. The business continuity is our biggest issue here, yeah. Overall, when we're talking about companies, we have to make sure that business continuity is there. There are some infamous examples of companies who went completely out of business because they never managed to implement their business continuity plan. They catastrophically didn't make it, bringing the business back to life because of simple errors or someone neglected something, didn't look at something, didn't check the right backup or did not store the backup at the right place. And all this stabilizes the performance of the business, whatever the business is, including providing the necessary resilience to ensure safety, security and continuous operation. This is the overall challenge here.

Amani: Yeah, that's really interesting. So as we are about to wrap up this discussion today, George, what would you say are the key points our listeners should remember about cloud computing, edge computing and PikeOS?

George: Yeah, cloud computing, scalable, flexible resources, computing resources from centralized data centers, it is ideal for tasks requiring computational power and large-scale data processing. Edge computing of the other end processes data closer to the source so it reduces latency and improves response times. It is perfect for real-time applications like IoT devices or autonomous vehicles. PikeOS can help in those directions as a robust, secure and real-time operating system it is, which means that it can enhance both cloud and edge computing environments. It offers strong partitioning separation, which is interference-free solution, other isolation, which means that one partition cannot interfere with the other. It offers virtualization of course, so it can be ideal for safety critical applications, ensuring reliability, security and compliance since it is fully certified for several different industry solutions. All those technologies can collectively improve performance, security and business continuity.

Amani: So now if we look toward the future, what trends should we keep an eye on in this area of cloud and edge computing? Are there any upcoming innovations or shifts or cybersecurity concerns that you think will impact how businesses use these technologies today?

George: Yeah, it is a topic on its own. Let me try to be brief. AI and machine learning for sure is going to be more and more challenging in the years to come. It is everywhere and it will keep on dominating almost all different aspects and applications of computer science applications. 5G will soon become 6G. It is offering widely demanded solutions. It is not mature yet if you ask me, but it will become necessary very soon because of AI, because of autonomous driving, because of drones and because of medical needs. So 5G, we should keep an eye on 5G as well, including IoT, industrial, medical, real-time. Cloud strategies will shift into hybrid and to multi-cloud. Why? Because there are many solutions aspects today where we want to combine cloud computing with that cloud. We need something in between or we need to shift from the one to the other depending on the solution. So over there we expect to see more and more things.

Security is a recurring topic and especially security for applications. Today we are heavily talking about quantum cryptography. I think this is coming very quickly and it will be very challenging for business to align, to certify applications and technologies with post-quantum cryptography. It needs a lot of work. In conjunction with AI of course, another AI topic is the Edge AI. More and more solutions. We work in such a European project called Rebecca. Right now where the goal is to bring AI into the Edge. Solutions like smart cameras, smart IoT devices are more and more in demand. The doorbell I mentioned before or the intrusion detection sensor at the gateway of a factory. Those are just simple examples.

Sustainability and green computing is something that the European Union is heavily investing and looking into it so we have to make sure that we find clever ways to adhere. Regulatory compliance I mentioned before GDPR and CRA. Those are two very hot topics. We have to make sure that all the solutions provided are aligned with legal demands, especially in EU where those things are very strict. And it's not easy to comply. A lot of things will change. For example, I'm looking at a lot of applications or programs or products that will go obsolete very soon because they will not be able to adhere to the CRA mandate of the European Union. Just a few trends that came up many more always having whatever has to do with security, real-time operations, medical and communications are all those topics that will keep on being at the center in the near future.

Amani: So on that note, we've reached the end of today's episode. Thank you so much, George. And to everyone tuning in, if you enjoyed this episode, don't forget to subscribe and share your thoughts with us. Stay tuned for more discussion on the technology shaping our world. Until then, take care and keep innovating.