How FB, Apple, and Google destroy tradition in the computer business

Andrey Chernyshev (Delta Solutions)

Today I wanted to talk about the interesting, as it seems. That, probably, more than fifty percent of us do not know well. We all know what OpenSource technologies are in software and, perhaps, some surprise for many of us are OpenSource technologies in hardware , in hardware. Today, I will try not to spend a lot of time on this, but tell the most informative about what kind of revolution is taking place and how it can affect us in general.

The trendsetters in this whole story are our big grandees. Ours, not in the “Russian” sense, but in the “well-known” sense. They are moving in this direction against their will. This is now and discuss.


')
As you know, the biggest costs for large web projects are the data center component, the component that costs them most in capital costs. And in operating costs, too. And new movements in the part of the server infrastructure are pushing the entire data center-building industry and, in general, the entire infrastructure industry to changes. And these changes are by and large revolutionary. Let's not be secret, this is true.

And the most important thing is what they struggle with. Those of you who have encountered infrastructure problems within your projects are all afraid and in general everybody is struggling with a certain vendor, with a certain lock on a particular equipment manufacturer. This is connected with the price component and not only with the price component.


The essence of this whole evolutionary revolution is to try to simplify the infrastructure. And the grants are fighting for the simplification of each piece of infrastructure data center. And easier - it is clear that cheaper in the end, well, more reliable.



This vendor-lok notorious does not just hold everyone by the throat. In Internet projects, the dependence is somewhat less than in the enterprise, but it also exists. And, most importantly, why we are fighting with him is the cost of infrastructure.

Of course, the grandees set themselves the goal and wondered: as much as possible, not by 3-5-10-15 percent, but significantly affect the cost of infrastructure.



They organized into a consortium called the Open Compute Project about which today, a little later, I want to say a few words. In our reality, in which we now live, there is some hysteria about import substitution, which moves not only in the minds of our officials, but, in principle, some kind of patriotism arises in our IT bodies. Because in the software we have learned to do a lot with our own hands and, in principle, we are proud of it. And now we are moving little by little to the fact that we in the gland will start to do something with our own hands. Import substitution is probably on these open standards - the standards that are written by the industry giants.



What to do? The most important thing is what to do? Trying to save on infrastructure quite a bit? Also one way out. But, as is known, all revolutionary breakthroughs are driven not by an improvement of 1-2 percent, but by some significant improvement, or a rethinking of what is now. That is, we are struggling with certain habits and with certain foundations. And this creates just revolutionary products.



The principles that allow you to compete with vendor-lock, compete with the price and, in general, fight, most importantly, for the effectiveness of those resources that are cheap in Russia, namely: electricity to a greater extent and, as was thought before, working strength. These principles allow us to look a little differently at the design of all information systems that we come up with, we make, we develop. And some postulates that go to those. The tasks for the new open platforms are well known to everyone:

• The first is hyperconvergence. That is, we stop separating storage and computing from each other. For us, software developers, this is a kind of universal, and Amazon has taught us to this, a certain resource.
• Next is the important thing. The main functionality is exactly what we need as software developers, because we can write everything else ourselves.
• That is, the minimal implementation of the functionality inside the hardware. All the rest we will write ourselves.
• A good path, a good understanding of where all this moves is also very important. And we want to influence ourselves on where the evolution of hardware will go, i.e. how to make our systems, our platforms that we develop, to make absolutely independent of the trend of the global market for the development of hardware. We want, on the contrary, hardware to evolve under our architectural solutions.
• And, of course, everything is completely open there, and we want to convince not only that Linux and other open source operating systems, but, in principle, would like to look at the BIOS. But this largely affects some specials. applications.

The main tasks themselves are formed from ideological trends, voiced earlier. In addition, it is important ultra-low energy consumption as a consequence of the condensing infrastructure. In the end, our electricity is also becoming more expensive, and this is a very large part of infrastructure costs.

What was born?



Two big ones came together ... We always like to say that there is Microsoft and something else. Unfortunately, there was no such big super grandee who could have run into it in the Linux world. We share, we can say, this great landscape of Canonical and RedHat among themselves. So, large consumers of infrastructure solutions, which, in principle, are coming up and deciding now for us, where to go the industry in this direction, are all of us beloved Facebook and, beloved, probably Microsoft.

And two giants in one project, as it happened by chance, implement two new interesting concepts. All of them take off under the OCP mushroom (Open Compute Project) and what is commonly called the Facebook Pure OCP (I’ll tell you a little about it now), and Microsoft is moving its own initiative - the Open Cloud Server.



Looks very different. In short, Facebook has always built the infrastructure for itself, including data centers. And the idea is to give the internal developers of Facebook some components, some bricks (bricks) from which any solution of any architecture can be assembled. Absolutely any. This gave birth to a certain unification within the components, and also opened up the possibility - first of all it is connected with the pricing structure - opened the possibility for third party-manufacturers to make compatible components with this infrastructure, with OCP.



The components are very simple: these are storage-modules, these are computing-nodes, these are micro-servers, these are network solutions and power management systems.

Within the allotted time, there is no possibility to talk about the details, about the technical implementation, if someone is interested, you can talk about it separately, but the most serious breakthrough that they made is a breakthrough in the efficiency of using electricity inside the rack (server rack). If it will be interesting, I will tell about it further.

Most importantly, these are all bricks, all of these components, they are completely independent of each other. Served separately, upgraded separately. Very comfortably.



What does Microsoft have? Microsoft moved its Open Computing Server as part of a project to deploy Azure to the whole world. Somewhere, where they did not fully believe in the markets, they used standard 19-inch infrastructure, or rented existing data centers, with existing racks and, as a result, there is a certain legacy mark in the standard, like an acquired trace, this is exactly the use standard 19-inch racks for server and storage management

There are disadvantages and advantages of this approach. To a greater degree, the density is very similar to Facebook, but not like this. There are problems with heat generation - struggling slowly. There are problems with electricity, but catching up with the use of special power supplies. At the same time, from the point of view of the cost-model, they did not turn out so badly.

In our opinion, to consider this branch, this fork real Open Computing-project is interesting if you are still building your projects on Windows and are planning to develop the infrastructure on Windows.



There are only three components: compute components, storage components and mixed components. And this is just one difference from Facebook. The network infrastructure in the Microsoft concept is not part of the project itself.



Two interpretations, two "Big" are struggling. Is this a technical war or a political one? This is actually a political war.

Some of us continue to work for a large company, some of us work for ourselves, but sometimes we find ourselves thinking that we want for our management, for our company, to be an indispensable, let's say, important link. This has a very serious effect on the things we do.

By the way, we posted in open access the video of the last five years of the conference of developers of high-loaded systems HighLoad ++ . Watch, learn, share and subscribe to the YouTube channel .

Perhaps this is not true in the first instance, but the people who move the two forks are driven mainly by this action. That is, they develop their individual forks and concepts solely in order to have a field for activity. Because in terms of goals, they are one.



Is it war or not war?

War! First of all, the war with Intel. She is very serious. No matter how much we love him, but, unfortunately, he currently has no competitors in the processor market. I will make a reservation: there are practically no competitors in the processor market.

And most importantly: it is very important that the lack of competition always, you know, stops a little progress.

The same ones, if we talk about Microsoft, they are “out of the race”, but Facebook, flirting with ARM some time ago, gave birth to a new project on the sidelines of Intel. And this project is now well known, it is no longer NDA'ny (NDA = Non Disclosure Agreement - non-disclosure agreement). This Broadwell-D is a new processor, which has a marketing name Xeon-D. This is a single of chip processor, which is specifically designed for Facebook's needs initially.

I confuse all the time: “hip-hop” or “hard-rock” Virtual Machine in PHP is used by J, but the point is that all Facebook frontends work on PHP, and, in principle, these are quite heavy frontends. And the processors of small, small-core, with low-power cores, was always not enough to twist this story. And to turn it on dual-processor Xeon's is quite expensive in terms of cost and efficiency. And thus, in competition with ARM, a new processor was born, which Facebook, thanks to this political war and the struggle of standards, gave birth to, in general, for the whole world. And now, I know, I now talked here with several guys, even in projects on the frontend, Xeon-D has been used to its full height, have grown out of Atom pants for a long time.

These are the changes that in fact really influenced our lives. And we see this, as Open Source Hardware Computing itself affects the whole of our lives.

What happens next? What are the benefits in the end? New products are being born - what do we get from this? Can we use it somehow? Of course we can.

New stories with Cavium, with new processors designed, designed mostly for storage-applications, which we begin to use everything little by little. This is the same SIF, for example, or distributed file systems. They are demanding on the cores. And, in general, these are processors, and this silicone, and this iron, which is being developed, including printed circuit boards, servers, and so on, is being developed thanks to these guys, who are actually fighting with each other.

What is the signal of all this? The signal is as follows: a large customer is now able to influence the product line from which our infrastructure is built. On the entire product line. If earlier, be it Hewlett-Packard, IBM, they offered the market a product, formed on the basis of some of their own opinions, that will be sold in one or another SMB, or in one or another Internet project, or in one or another large enterprise customer. Now the situation, a couple of years ago, began to unfold in a completely different area. Large customers determine what the infrastructure should look like, and this is done not without the help of colleagues involved, in fact, with the software infrastructure.

Do we need to follow this? I think it is necessary.

How do we know about this about everything? In our business, we are developing a real-time platform for telecommunications operators. Recently, when the ruble was divided in half, or the dollar was multiplied by two - as everyone likes to call - customers want to consider everything in rubles. They want to consider first of all in the context of the whole project: hardware plus software, all in rubles. The ruble has fallen in price, which means that in the complex project, part of the infrastructure has doubled. That is, in general there is no money left for software and work. And it made us look at this story. Since life forced to write software, which, in principle, is not very demanding of infrastructure in terms of redundancy (redundancy, duplication in order to increase reliability). Perhaps this is more the same thing as you do, the same story.

Looking at the "struggle of the titans" in detail, we saw that the new standards provide an opportunity not only to develop their software, but also hardware.

If there is a moment - look, they are interested on the consortium website, new licenses have been uploaded, with the help of which, after reading them carefully - there are pitfalls - you can develop your own hardware within the existing constructs. I'm not talking about any processors "Baikal", what drives our state, with the support of "Elbrus" or something else. We are talking about constructs required for your applications.

Not so long ago, the guys from MAIL.RU, discussing the optimization of their mail, clashed and talked about a very interesting topic - the processors are becoming more powerful, the cores are more powerful, and the cold storage, as it was cold, now becomes even colder. Because from the Nativity of Christ we keep all our photos there. And thus the requirement for the number of spindles that need to be added to the same or slightly increased computing resource of the processor, it increases. And while the industry is not very running in this direction, since these custom applications are not so many.

Nothing complicated: Inventor, Solid Works, a couple of good engineers who know how to work with metal, a part of motherboards that are designed by industry giants and have normal reliability parameters - and here you have a customized device in your form factor.



"Big" what are they doing?

Formally, everyone joined this consortium in order to join, to be somewhere nearby. But HP moves, for example, with FoxConn, where FoxConn produces hardware, HP is trying to sell them under its own brand.

Cisco uses the same Braadcom Trident 2 chipset in Nexus switches, which is used in Bare-Metall switches in Facebook, Amazon, Apple iCloud with Open Network Linux. And, in principle, this is probably the dream of any developer - to try load balancing on the switch, not on the servers, and do it with ASICs. Here you are! At the same time, Cisco began to sell its switches without software. Do you think they have become cheaper from this? Of course not.

Hitachi, Ericsson, Nokia and others continue to buy ready-made OCP components, but they believe in this initiative, as it increases competition.

But in the end, why did Facebook start this business? Price! Remove the marketing component, remove the brand component from the cost of the final hardware. Because Intel is alone everywhere, Broadcom chipsets are everywhere alone, Mellanox chipsets are alone everywhere, five companies in the world make memory, produce memory chips.

There are manufacturers that are large, the main producer for Facebook is QCT - the entire range produces compatible equipment and, in principle, I cannot say that for all occasions, but covering part, including these tricky MAIL.RUs problems, I think that not only they have such problems there. In general, the idea is to find a good supplier, if you are not yet ripe for your own decisions.



Ready to discuss if anyone is interested in all this OCP's story, ready to discuss it in detail, tell what benefits it can give to your specific project, how to deal with the cost of infrastructure and how to bargain with your Cloud Service Provider about the cost. even rented infrastructure from him. Come, we can discuss separately.

That's all I wanted to say.



By the way, I'm sorry, these machines with Xeon-D processors, which are called the “Yosemite” project, they arrived to us not so long ago. Facebook had an embargo on selling them to the side, so it was overcome. You can come, touch it on the stand, look - quite an interesting interesting thing under the frontend.

Thank!

QUESTIONS AND ANSWERS

Question: You mainly talked about the opposition of Microsoft and Facebook, and in the title of the Facebook report, yes, but there is Apple and Google. Somehow about them could? Already Apple, it would seem, it is proprietary-proprietary. And what is Google's role?

Answer: Actually, in March of this year (2016), the terrible happened. Google was a supporter of non-participation in all these wonderful stories, but in March they joined the OCP project, they are already there, the documents are signed. Now there is a battle, they are fighting with Facebook'ovskimi engineers on the topic of power. That is, they are satisfied with and constructive, satisfied with the whole approach.

If you delve a little bit in the technique, literally two minutes. The entire history of OCP says that the blade basket, where the infrastructure is inserted (not only servers, but also storage, and network components), the blade basket is rack. Those. The rack is a blade basket and centralized power is on the rack. Here, in the Facebook version, this is 12 volts, while Google’s own projects used 48 volts, arguing that a smaller cross-section of copper conductors is necessary in order to transmit the same current. And now in the new consortium, I participate there, I see the correspondence - in the new standard of the rack they add the ability to use 48 volts as power.

They also act as an initiator to add a special bus to the rack in order to be able to manage the components among themselves. Because now in the concept of each component is a separate island.

Google is there, it has not gone anywhere. From this we indirectly conclude that the Microsoft story is probably “yes”, but the slippers are put in a different basket.

Q: Ok, and Apple?

Answer: The guys from Apple are one of the serious participants of this consortium and the last, the largest data center, which they have built for Apple Music, it is completely OCP's, it is the Facebook direction, it is completely OCP's. They are involved in management, are involved in the management part, and they are more a consumer rather than a contributor.

They are the most serious proponent of dragging ARM processors into the fight, as their history with the Power PC that they had before, c migration, respectively, to Intel and the behavior of Intel, which they now have. We understand that everyone conquers the market with a cheap price, and then they have to recover the margin. They use this argument and take it to the table in order to get good financial conditions from the guys. Because in reality, Intel has no competition.

There is OpenPOWER, which is moving slowly and, in principle, Rack Space is one of the big contributors, the only one who uses OpenPOWER under KVMs. They invest a lot in it, but IBM is cunning. That OpenPOWER, which is available now, it is only with DDR3-memory, it essentially loses DDR4, we know. There are performance issues, but Rack Space believes in them.

Q: And the long-term this philosophy, it does not pick the bitches on which Apple sits? Because after all, talking about commoditization is, and Apple sells the same pieces of iron at twice the price for some unique brand.

Answer: Now we are not talking about Open Source Hardware on the desktop, not about Open Source Hardware in our pocket, but we are talking about Open Source Hardware inside the server farm.

Question: I am talking about long-term philosophy. If the market turns somewhere, then everyone starts to take it and ask questions about MacBooks too, probably.

Answer: This is a different market - MacBook. If we talk about the server economy, of course. There is another ideological piece of history, at this closed event one of the vice-presidents of Intel said an important thing. In 2017, they predict 60% of server processors to sell to Cloud Service Providers.

Can you imagine what it is? This means, and now only 30% is on sale, which means that according to their forecasts, and they, in principle, are smart people, according to their forecasts, the market of server processors will drop dramatically for everyone else, except Cloud Service Providers. This is the answer, there will be a crazy war just for this market.

This report is a transcript of one of the best speeches at a professional conference of developers of highload systems Highload ++ , specifically, the section “Hardware, Infrastructure”.

The company Delta Soultions , which is headed by Andrei, is a partner of the festival " Russian Internet Technology " and will bring new pieces of iron for the exhibition, which can be felt and studied.