I'm really worried about Redhat. IBM is currently a sinking ship, they've laid off parts of their workforce, they're failing to capture the cloud market, and the Redhat buy seems to be a ploy to maintain relevance. I'm seriously anxious about Redhat being drug under during IBM's death throes.
Redhat isn't going to save IBM. They sold because they had an unsustainable business, and Openshift is still a risky play to run a stable business on. The container space is starting to get awfully crowded, and cloud providers can provide better tailored experiences and skip containers altogether with serverless.
As a long time taker of Redhat certs, I hope they pull through and still look something like the Redhat I have loved and known.
I've been keeping track of the various Kubernetes services and agree the market is massively crowded right now. AWS, Google and Azure dominate the cloud market. IBM is a distant joint last place by most metrics.
OpenShift won't compete with the cloud offerings in the long run. Redhat are currently competing with On-Prem offerings and trying to make a case for 'hybrid'.
Even in this market OpenShift is competing against at least 15 alternatives. For anyone interested I've been crowd sourcing data for each and trying to work out key differentiators.
OpenShift is definitely differentiated in terms of security. They also have network and storage features that are arguably valuable to enterprise customers.
Google GKE On-Prem is out soon. I never see Azure Stack mentioned so maybe Microsoft will look into a Docker acquisition to boost their enterprise footprint.
The cloud seems like the bigger growth area though so perhaps this On-Prem container platform land grab will be short lived.
> I never see Azure Stack mentioned so maybe Microsoft will look into a Docker acquisition to boost their enterprise footprint.
The companies using Azure Stack today are some of the least likely to be parroting that fact. Microsoft's on-prem is pretty well entrenched, particularly in "boring enterprise" where pragmatism beats idealism, are quiet "do the work, don't leak the work" more than proudly proclaim to be chasing the trends.
That said, Azure Stack (and Azure) already have strong Docker support, without an acquisition to date.
More interestingly, in my opinion, though, is that right now Azure Functions is the only "serverless" runtime with a strong open source story, and equally strong on-prem (both as part of Azure Stack and, increasingly from what I hear, standalone) as well as Cloud support.
I'd be surprised if Azure Stack sees big adoption any time soon; they aren't just letting you run some Azure services on existing hardware... the way they're doing it is to really take the "cloud in my datacenter" concept to the extreme:
* Subscription pricing like a cloud service, even though it's on-premise.
* Only runs on certified hardware appliances... You're looking at an investment of at least $300,000 for a base configuration.
So, OK, I get it, you're getting a private cloud and you're not supposed to worry about hardware anymore, but it's probably too big a mental shift for most traditional enterprises. And people with that kind of budget are already locked into a VMWare infrastructure for the next couple of years.
> That said, Azure Stack (and Azure) already have strong Docker support, without an acquisition to date.
Maybe I'm confused about what you mean by "Azure Stack", but there was most certainly an acquisition[1] related to Azure's docker support, the Deis team (now part of Azure team) were some of the earliest contributors to the Helm project, and this buyout was widely hailed from what I can remember as Microsoft's first big investment in Kubernetes.
Many of those people remained with the Azure team, and none of them work on Deis today. So, instead, they are driving the development of tools like Draft, AKS, OSBA, and Virtual Kubelet.
And of course MS also made a high-profile acquisition of GitHub, just not sure that one is relevant to Azure Stack.
Interesting, I would say that it's fortunate nobody can buy Kubernetes (because it only is composed of acceptance and compliance with a conformance suite, there are dozens of separate implementations and they may all be Kubernetes.)
To be honest that's a great feature and one reason why I am not terribly scared about consequences of any news about a RedHat acquisition; that being so even though I do follow OpenShift with some interest.
I'm sure that Docker is also a target for acquisitions by these bigger companies but while we're on the subject thanks to Kubernetes and the CRI, it's pretty darn unlikely that anyone is ever going to totally corner the market on Linux containers, at least any time in the foreseeable future.
Fascinating. The choices made in the criteria in that report are alluded to but not explicit, so of course we must resort to mindreading. My best guess is because Azure Stack is more than just a "container platform", and the "container platform" for Azure Stack essentially is "just" Docker, so Azure Stack usage in this case would be rolled up in the strong position you see for Docker here.
It would be great to see Forrester more forthright on such methodology, however, and this may be a good reminder that Forrester here is not a transparent, peer reviewed journal and doesn't have to show their homework.
Docker, as mentioned in the Forrester report is "Docker Enterprise 2.0" and as the text mentions it's a Kubernetes platform with Swarm and Kubernetes orchestrators.
AzureStack presumably runs AKS, the same as the Kubernetes service that runs in Azure, which is totally different.
The market size can't be attributed in any way to Azure Stack which is why I was surprised you mentioned it was heavily used. Gartner talks to a wide range of companies usually.
> "We did not evaluate native public cloud container or Kubernetes services."
In which case AKS doesn't show up because the report intentionally excluded it.
Again, I don't understand anything about the methodology of this report, it's very clear it's designed to present some bias, but there isn't enough transparency on what/why/where, and every time I skim it I come up with other questions as to the validity of this report towards any market size determinations whatsoever.
The report is from Forrester. I hardly ever read them but Docker has it posted all over its website so I was curious.
Forrester and Gartner seem too high profile to be biased. This one isn't sponsored so the researchers will have had free reign over what was added. If Azure Stack is really in use by many companies I'd be surprised if they missed it.
Strictly speaking, Cloud Foundry is not Pivotal's, it belongs to the Cloud Foundry Foundation. We're the leading contributor and we sell the leading commercial distribution, but it's still an open source project.
We compete with Red Hat in on-prem and hybrid scenarios but we're also cooperating with Red Hat, Google, IBM and SAP on Knative.
While all of these solutions are maturing at a rapid pace, the problem of cloud is still cultural for most enterprise companies vs. tech stack. Operational discipline to make cloud a business driver is still an unmet need on every large enterprise side and it has lot to do with the non technical items, culture, incentives, organizational models
If you really want to get into the details, Walmart has their own hybrid solution which has been open-source for quite a while, OneOps. It's a terrible product, but somehow they manage to use it.
Seems like Redhat is to IBM as a new puppy is to a failing relationship...
However, I am not optimistic that a Glorious Future will be the outcome from the RedHat acquisition;
While at Mirantis, I was assisting in RedHat's implementation of OpenStack with Intel/IBM and it was a nightmare.
OpenStack, at that time was very difficult to implement with smooth success (much like a JBOSS / WebObjects / Tomcat nightmare implementations back in the early days) -- and the BigCo process-procedure-competency-silos that goes along with it...
I can only expect that adding more layers of abstraction and ambiguity of ownership will make for headaches.
Here is to hoping that my pessimistic outlook is proven wrong.
OpenStack is still a nightmare, no matter what distribution.
OpenShift, on the other hand, is a breath of fresh air (k8s in general is - the quality is so much higher). Things just work and the docs don't conveniently leave out all the important details.
Not really. Unless you haven't tried it recently, there are many great options. Kubeadm is painless for simple clusters, and kubespray is easy if you know ansible. Of course, both assume some level of sysadmin experience, since it's a tool to manage large clusters.
You-know-what, I owe myself and everyone involved with that project a little more info:
I was the lead PM responsible for managing the implementation of the OpenStack on RedHat+IBM with Intel TXT being the premise of the implementation such that RH could go back and replicate into their own thing...
It was a nightmare project, but on no part, individually of RH, Intel or IBM or Mirantis...
Openstack just was NOT READY for production secure workloads.
The project was just a bunch of people trying to do their best with shitty tools...
But, openstack/shift/the-current-marketplace wouldnt be without the sacrifice of crap on the battlefield.
So, at the end of the day, just hope that all your failures are something that others can trample to make-the-world-a-better-place...
OpenStack screams “enterprise” and “architects” at the top of its lungs and I for one want nothing to do with that crap. I know exactly what it is just by looking at it - a hobbled together patchwork soup of poorly documented and half-abandoned open source middleware that is meant to pay consultants’ salaries and nothing more. Definitely not meant to solve real problems, that’s for damn sure.
Who is fooled by this anymore, really? Non-technical middle managers who read a bunch of marketing materials?
IBM isn't sinking. It is quite large, nearly 400,000 employees. If they laid off 10% of their workforce it would be like a bit more than 1 and a half Facebooks ceased to exist but it would be a drop in the bucket at IBM.
What they are is slow moving and hidebound. I think that at some level they understand that, and at other levels they don't. They have an amazing asset in their labs where they can put 1,000 of some of the smartest people you might ever hope to meet on to a problem pretty much instantly.
In my short time with them post acquisition I found a company that was very slowly losing its calcified husk and becoming more agile. It had some blind spots but when you replace enough people that can be fixed. As Gerstner showed in the 90's, as Satella is showing at Microsoft, what Welch showed at GE. Put the right leadership team in place and you can do amazing things with that power.
It's all contractors being sub contracted to other companies. If they laid off 10% of their workers overnight it wouldn't matter because it's the normal cycle of contracting.
There can be no growth because contractors don't scale. They bring a fix percentage of their day rate.
There could be some products with great people lost in a corner of IBM. However, they will remain a negligible fraction of the company and they won't have the power to turn it around.
If AWS/Microsoft makes a new product or service, they can roll out instantly everywhere and capture clients. If IBM does something new, they couldn't begin to sell it, they'd sell you a contractor.
> There can be no growth because contractors don't scale.
Except of course that when you employ 300K of them it does scale. The mistake many software people make is that only software scales, but IBM (and others) has successfully built a scalable model around consulting.
> Put the right leadership team in place and you can do amazing things with that power.
Sure, but that's the thing: IBM doesn't have the right leadership team in place, and hasn't since Gerstner left. It shows no sign of having learnt from its mistakes.
Fundamentally, IBM is a sales company, not a technical company. For a long, long time its marketing department was able to sell the fiction that it was a technical company, and use that reputation to sell, but folks realise now that it's just not.
As a recent IBM Hire...one of the first in about 30 years for my role and definitely the first for my region ever I am stoked to work for IBM. Me being hired quite literally wiped about 5 companies contracts out of my region for certain hardware roles...which they deserved because they provided terrible service. IBM is certainly becoming more agile...the one thing I am learning about my role within the company...is that change is a constant for the company.
Businesses this big and old don't "sink" unless there is utter contempt for obvious changes in market conditions by the entire board, for a sustained amount of time. Even with utter contempt, it would take an enormous amount of time for them to implode. IBM isn't that stupid and they obviously understand something isn't working.
They're also roughly 100 years old (depending on definitions). They're reinvented and reshaped their business many times. "Two new broom handles, two new brushes" but it's still the same company, as they say. I'm afraid I find it ludicrous that IBM would actually end up in anything resembling financial failure.
> They're also roughly 100 years old (depending on definitions). They're reinvented and reshaped their business many times. "Two new broom handles, two new brushes" but it's still the same company, as they say. I'm afraid I find it ludicrous that IBM would actually end up in anything resembling financial failure.
People were saying the same thing of Sears 10 years ago...
Sears falls into the "utter contempt" category, along with utter incompetence. It was obvious to every outside observer their model was busted. It was obvious how bad their stores were to anyone who went there to try to buy something. There was nothing sudden about it, really. I think many could see this happening more than 10 years ago.
Also, unlike IBM, Sears has very little of worth which would uniquely identify it as "Sears" after asset sale. It's all fungible land, rent, concrete store walls, and staff. They have many competitors with practically identical overlap in business. IBM is not like at all like this.
It's all fungible land, rent, concrete store walls, and staff. They have many competitors with practically identical overlap in business. IBM is not like at all like this.
Ermm yes it is. What do they do that HPE, DXC, Accenture, Crapita, Fujitsu, TCS, Infosys, Wipro yadda yadda don’t? All IBM had left is the rapidly tarnishing prestige of their once mighty brand.
IBM's customers are the same as Microsoft's customers (at least the high profit ones). Very large enterprise companies locked into contracts worth millions and 100s of millions of dollars. They move slow like icebergs and carry as much weight.
They don't worry about new customers when the largest companies in the world have gone through months or years to get them as preferred tier one vendors.
Sears on the other hand did not serve large Enterprises and were generating revenue mostly from retail. Consumer vs. Enterprise is like Apples and Oranges guys.
Hum... I work in a large MS customer. We are a happy customer for some measure. We are also an Oracle customer, and happier than I would expected to be possible if I didn't know about there.
I just can't imagine anything we could hire from IBM.
I’m not sure that’s true. In the days of film the average consumer shot <1 roll per year. Whereas a professional photographer would shoot a dozen or more rolls per day. Their biggest customers would have been newspapers and studios buying 10,000 rolls at a time on a regular basis
Would you mind entertaining a naive question? What makes redhat worth it? What would make a company choose redhat over Ubuntu (or any flavor of Debian) in 2018?
RHEL's quality assurance and documentation is top-notch. Ubuntu has been improving, but it's not anywhere as good (which makes sense - Red Hat has much more people working on it).
Support is expensive, but worth it. It's not unheard of to get custom kernel patches mere days after you reported an issue.
Well, that and the ecosystem of enterprise software which refuse to run on anything but RHEL.
And Ubuntu's failure to establish their home-grown stuff: Upstart, Unity, Mir, Ubuntu One, Ubuntu Phone, Juju...
That being said, 18.04 is a lot better than 12.04 or 14.04. I'm a Red Hat fan, but I hope Canonical succeeds - the more independent companies there are, the better.
I've never seen Ubuntu used in a large enterprise. Redhat dominates the server market. They offer better certification, support, and a package repository geared towards stability and backporting. The seperate Fedora (new a shiny) from RHEL (old and stable).
Ubuntu doesn't really have that distinction. Shuttleworth is also kind of a character, and it's hard to rely on him keeping Ubuntu fixed on a particular use case. It's been desktop, then mobile/desktop convergence, and now containers and Kuberentes. Think startup vs big business.
Ubuntu is the default base image for Cloud Foundry, so in fact it's pretty widely embedded in enterprise at this point. But it's an implementation detail, rather than a separate line item on the invoice.
Google originally used Red Hat Linux (not RHEL) but migrated to Debian. They are using a custom version of all their stuff so its probably not really even Debian at this point.
The docs are just that much better. The server packages are much more reliable. If you want specific software versions, and you're not on docker, you get those versions. It's a breeze to sysadmin a centos machine. An Ubuntu machine is a pile of nightmares.
(Also, very few people actually run RHEL with the support package, most run CentOS, which is patch-compatible but has all the trademarks removed.)
You will occasionally run into problems requiring much deeper systems expertise than in a standard Ops or SWE skillset. Red Hat is a practical and economical alternative to hiring kernel hackers yourself.
Most of my experience with IBM has been pretty reliable:
1. They bring the A players to the sales process
2. They bring the B players to manage the account (technically and commercially)
3. They bring the C/D players to do the actual work.
Wondering if I'm an outlier or if this is a standard mode of operation for them? I wonder how this plays in with their OSS strategy? Who works on things like OpenShift?
They bring the A++ lawyers to tell you that not having database corrupted / application crash / buttons click fail wasn't specified in the contract and that in their opinion the software works as agreed.
Seeing the IBM letters still leaves a bad taste in my mouth.
Whenever I meet with any of the big(ish) players to to assess a product I spesifically ask them to send engineers instead of sellers.
If they send sellers after that we usually just ask them to leave and send the engineers. I've seen quite a few pouts but it's impossible to get accurate product information out a seller, they just tell you what you want to hear.
I've never been in contact with IBM before but I'm not surprised about your experience.
You can usually negotiate that after you've confirmed that the product will actually do what you need it to do.
When I took sales calls, I would usually sit through the intro pitch from the salesperson, and if it seemed in the ballpark, I would cut the call short and ask to be put in touch with product/engineering by email. And if I liked what I saw, I'd go back to the salesperson for pricing, etc.
First I want to know what the product actually does, then I can worry about prices and contracts.
They are usually the least part of the problem.
The biggest issue I've had in the past is that I ask for a feature and the salesman say "no problem" and when it's time to implement what I asked for is a custom feature that they can build for €100k.
The observation about getting C players to do the work is part of the actual strategy for succeeding in their market. The idea is that there are C players who are supported by B players who are supported by A players at HQ so that when things aren't going well the problems are escalated in order to either make the account go well or fix it after something has gone wrong. It is a model that actually works - though it can be a rocky path sometimes.
If you hire an independent A player as a consultant instead, they have no big organization with the resources of IBM. I've competed with IBM, and the saying of "you don't get fired for choosing IBM" is still more true for them than for anyone else.
IBM can still bring a level of competency which far exceeds anything Red Hat alone can do. Several well-known and successful companies where I've worked considered their biggest competitor to be IBM - though they didn't say that publicly since they didn't want to further legitimize the competition. They were already losing too many sales to IBM.
I worked for an independent team of A players (that ended up being bought by a D player company), and it was pulling teeth to get deals. We were better, everyone knew we were better that ever spoke to any of us, and them.
But none of that mattered. All that ever mattered was the CIO, and what company he wanted. Then he would write up a rule system where X company had to meet Y requirements that only his preferred company could meet.
We still got more work than we could handle, but people always want A players on the ground, but even at the exact same cost, the upper level management are going to make decisions based on things that have nothing to do with success.
> The idea is that there are C players who are supported by B players who are supported by A players at HQ
I agree it's a strategy, just not a very good one IMHO. I'd prefer to see A players at all levels, just at different levels of their career. Junior level A players can still contribute.
Your assumption is that a company with 300,000 employees would be able to only hire A players. The larger the company, the more it will naturally HAVE to trend towards average. Most successful large companies are successful because they know how to organize their company so that average employees are productive.
I'd be surprised if you could find enough A players to staff the model at the present payscale. Sure, you can argue that a more successful entity can generate more profits, but there's diminishing returns to that effect. Furthermore, there's going to be a lot of egos in the room if all you have are A players, and that's not necessarily sustainable in the long term, either.
A more successful entirely does not generate more profit in consulting body shops. The only thing that generate more profit is more consultants billed for more hours.
You still hear the old adage “nobody ever got fired for choosing IBM” in many large enterprise software sales negotiations, even if the veracity of this statement is increasingly dubious in 2018.
And likely there are many counterexamples proving that people do get fired for choosing IBM but companies and governments typically do not go out of their way to advertise their failures.
Not in my experience - at my last job we had multiple meetings with them to try and sell us one of their products (CIO was ex IBM and was pushing us to use it) and everyone technical who was involved on our side was profoundly unimpressed.
Was also a bit bizarre that about 6 of them came along to a meeting with 4 of us - we never did find out what most of them did.
> everyone technical who was involved on our side was profoundly unimpressed
That's your "mistake" (from their part). They only know how to sell to top-level oxygen-deprived-brains that are impressed by technical fancy words but couldn't turn their notebooks wifi off and on again to save their lives
>Was also a bit bizarre that about 6 of them came along to a meeting with 4 of us
Ever seen that scene in Halt and Catch Fire where IBM comes to meet with the small company and a hundred IBMers flood into the office? That's the most accurate scene I've ever seen in any movie or TV show ever.
> Was also a bit bizarre that about 6 of them came along to a meeting with 4 of us
Sounds like that's a typical mode of operation, seen that several times, too. So maybe the quality of the sales team is the variable. So if your sales team is C players, do they bring the... E? ... players to project?
It was actually quite an uncomfortable demo as when they finished they genuinely expected us to be impressed - what they had demonstrated would have taken a few moments by someone competent in SQL to duplicate.
I think they might have been used to giving demos to not technical folks rather than people who were actually fairly competent and inclined to scepticism.
I'd been putting Dell/Netapp hyperconvergence appliances into accounts for 9+ months when I wound up at an IBM dog and pony show where they demoed an early version of their own box and acted like it was the first time anyone had ever conceived of such an idea.
Pretty standard for consultants in general. I worked for a contractor where we did proof of concept after proof of concept for Kubernetes deployments and the people working on the proposals were literally the only people at the company that understood Kubernetes. It was a pure bait and switch. We had nobody that could actually implement what we were selling.
Very rarely does sales get dinged if they make a big deal for something that doesn't exist yet, or even something which is completely impossible. They cash their bonus checks and move on to the next mark, tossing a big mess into the laps of the people that have to implement. In a lot of the bigger shops, people move around so often that they aren't anywhere near by the time the shoe finally drops.
Wow!!! This sums up the culture of IBM as confirmed many replies to the above comment.
> 1. They bring the A players to the sales process, .. C/D players to write code.
I was wondering why IBM failed to make great progress with WATSON project which was started years ahead of others in ML/DL filed.
Now I know the answer, internal IBM projects are developed with same A/B/C people of the hierarchy, where C players are people who write machine Learning/Deep Learning WATSON code.
Since Microsoft lost consumer space by losing windows phone, it's focus is ONLY Enterprise market which is confirmed by LinkedIn purchase.
Slowly but surely Microsoft will eat IBM's "Enterprise lunch" eventually. (sharing part of lunch with AWS & Google Cloud)
I don't think it's fair to characterize the Watson developers as C players. At the time, developing a natural language query interface to a web-scale unstructured database was a pretty cool technical achievement.
The problem was that Watson was a (largely successful) publicity stunt. In that sense, it falls in the category of sales and IMO is a top notch, A-player, demonstration of hype. It wasn't designed from the start to solve real problems, so it was always going to be hamstrung as a real product, regardless of how good the underlying engineering was.
* Edit to add a link to the papers they published -- https://researcher.watson.ibm.com/researcher/view_group_pubs.... -- which I remember skimming at the time and feeling like they were on par with any other typical ML paper. In general, research departments are run with an entirely different culture than engineering.
Watson represents the end game of IBM's retreat from selling solid (literally) products.
When I worked at IBM eons ago, products were very tangible, and often very physical in form, and there was no doubt whatsoever about what they did. Each came with beautiful bound folders for Africa describing functionality and usage. As a Systems Engineer, a part of one's job was being able to drive the configurator and other tools that controlled updates and upgrades and ensured products interoperated correctly.
Watson, it appears, is more the sort of snake oil that one would associate with a flakey, fast moving startup that can promise anything and then somehow make it sort of happen for the early adopters. Its hard to pin down exactly what is IS - which is kind of the point when using it as a sales tool.
Watson is a brand name, that's all. If you want to dig under the covers what most people think of as Watson, i.e. the sentient being kept locked in the basement it is just marketing. The actual "Watson" services are competent implementations of Speech to Text, Text to Speech, Natural Language Processing, Visual Recognition, and open ended classical statistics and neural network based model generation/execution. It's no more snake oil than Google AI or Microsoft AI.
That's how all the big players in this area work. And the C/D players usually are in a foreign country where it's even hard to get hold of them during the day.
>>They bring the C/D players to do the actual work.
That is probably true for any company ever. Unless of course you are full of cash, overflowing to a point, where you have ginormous amounts of time and money to waste interviewing people, reject them and offer whom ever you want top of the market salaries.
Any company can probably do this at their prime. Other people have to hire on average competence if they want to scale. And that works fine most of the time.
> one of the most important takeaways from the Red Hat acquisition is the admission that IBM’s public cloud efforts are effectively dead
> while IBM will certainly be happy to have the company’s cash-generating RHEL subscription business, the real prize is Openshift, a software suite for building and managing Kubernetes containers
> IBM is betting it can again provide the solution [to cloud lock-in], combining with Red Hat to build products that will seamlessly bridge private data centers and all of the public clouds
To sum up - IBM lost the hardware cloud wars and is making big bets on cloud software. The author doesn't answer the key question posed at the end - does the problem of cloud vendor lock-in actually exist (and the implied question - if it does, how profitable is it)? I suppose that's the info he wants you to pay for ;)
Where I work the answer is most certainly a YES. We have so much stuff using non-portable cloud offerings that switching would be a nightmare. (Step Functions, Lambda, SQS) Sure we could switch
to replacements but the cost and time would be prohibitive.
At some point most large companies will play their hand and raise prices to more closely match the cost of switching off the platform. (i.e. the MO of Oracle/IBM) It is really more a matter of when, not if. The problem with doing this is customers wise up this strategy and quit using your products for most new work. Once this happens, the company's best option for growth is to buy new users via acquisition.
I don't see Google allowing this to happen, they will try to offer lower prices/extras because they need to make up for lost time and market share to leaders AWS and Microsoft who are quickly pulling away as the 2 leaders in the market.
Google has enough resources to accommodate any medium/large enterprises that want to switch through price and onboarding incentives. Not to mention they employ many of the world's Kubernetes developers, experts, and maintainers.
The holy trinity has enough cash and willpower to keep a monopoly/monopolistic pricing from happening and there's enough business for them outside of the cloud to encourage them not to collude IMO.
Not to say hybrid cloud architectures are unnecessary - they most certainly are good, and the way the industry is moving, I just don't think there will be a huge pain point over monopolistic pricing/actions.
Sure it might be cheaper to use a competitor but if it will cost us 10M to migrate and test the working production workflow then it would have to be 10M cheaper to switch.
This gives our current provider a strong incentive to charge us a bit extra. Say 1/4 the price of the migration. Nobody wants to start a high risk migration that would take 4 years to pay for itself. Easier to just keep paying 2.5M extra. Safer that way and it isn't their money anyway.
> At some point most large companies will play their hand and raise prices to more closely match the cost of switching off the platform. (i.e. the MO of Oracle/IBM)
I don't see AWS doing this. They've continuously implemented technology improvements that have allowed them to reduce costs for customers 67 times in the last 10 years. The price of cloud services always goes down, not up, over time, as you're able to realize the benefits of improvements in compute power for the same cost (Moore's Law) without upgrading hardware.
Disclaimer: I work for AWS, but my opinions are my own.
You are thinking like a software engineer, not a product manager or owner of capital.
If any of these companies is able to extract monopoly profits without fear of competition undercutting them or government/legal intervention, they will.
Right now all 3 are competing fiercely to win the hardware rental market, but this may not be the case in the future once the market matures, and they may raise prices to match leaving costs (as parent comment pointed out) if they can lock customers in.
The amount of implicit absurdity behind your words is astonishing. Basically outside die-hard Marxists nobody takes such claims seriously, because it implies lack of competition.
This makes me scratch my head. I heard Power9 is getting more and more traction in some clouds, notably Google's.
Maybe IBM lost the "bog-standard VM-based cloud" wars, where the "standard" is what AWS / GCP / Linode / DO, etc offer. But that space is indeed crowded; doing something where you have little competition makes a lot of sense. (I wonder if they'll be able to.)
IBM bought SoftLayer 5 years ago, and was unable to capitalize on their niche (bare metal). Instead they decided to double down on provisioning VMs, which SoftLayer was never that great at. The set of features IBM's VM's provides compared to any other modern cloud is terrible (headed backwards in Gartner charts). It's barely getting any better, meanwhile Google/AWS/Azure are gobbing on features.
IBM effectively failed in their integration of SoftLayer. They bought SoftLayer but failed to modernize them. Why will RedHat be any different?
Take a look at what's left of SoftLayer today. Most folks who had the necessary skills to keep their ancient php platform relevant have left a long time ago.
I'm sure IBM is on their 5th iteration of their "next gen" VM provisioning system by now, due out any quarter...
Have they fixed their scheduler? For 5 years nothing has improved in the process of launching VMs. Sometimes it just barfs and goes to hell, requiring contacting support to terminate provisioning.
No joke, at one of the previous places we’ve implemented a task which would raise a support ticket via API if provisioning step was hanging for longer than x minutes. This is the case for years!
According to some people from IBM, the reason why IBM purchased SoftLayer was so they could tell their customers who were looking to move to the cloud „you want cloud, we have cloud” ... and keep the hefty support contract runing.
More fatally than that, perhaps: IBM's business model for half-a-century at least including renting time on hardware at as close to commodity costs as IBM could push them (and still make a very healthy profit admittedly). The Cloud model was IBM's invention in a time where computers were expensive. IBM's failure to transition that very business model the company was built upon to a time where computers are plentiful, is like watching an Olympic swimmer forget how to swim in a lake.
SoftLayer is not competitive. It was undercut by Google Cloud since last year, while AWS is not that far off.
Once everyone offers a choice of 1 TB Virtual Machines for rent, SOftLayer will be completely out of the equation. They're strictly worse in services and pricing.
I think we'll see a trend of cloud fragmentation again, once there are more good cloud software stacks. Imagine something like OpenShift + OpenStack + Ceph that does not suck and can be controlled by something like Terraform and support various types of hardware. That's just a crude example I can come up with quickly to demonstrate the concept. Something like that will enable many more types of players in the market and people can use a mix of them or self host.
It will be good for IBM to be in the middle of this, since the first such stacks will probably come from a large player that doesn't have a dominant existing cloud service. It will also be good for IBM's hardware expertise.
I think this trend will come, companies are getting locked into AWS almost deeper than some companies got themselves nailed to IE6 + .Net at some point. This is fine when the benefits appears to outweigh the risks, until it doesn't.
I've never heard of power 9 getting any traction with Google, short of a press release 4 years ago by them saying IF power9 lived up to the hype, they would consider switching. Since then, I've seen no information that they actually moved any noticeable portion. It's certainly not available publicly. Even IBM didn't have power9 in their own public cloud last I checked.
All it said is they've deployed power. That could be 1 server or 1000. I think the fact still stands that of IBM doesn't have them in their cloud, who will?
Vendor lock-in is most definitely a huge thing. And being on the side that wants to avoid it and build portable solutions is never a winning argument. You're always the guy who's setting the perfect up as the enemy of the "works now."
I seriously doubt that IBM can execute on this plan based on my limited experience with them. But if they could, it would be a winning move.
Good insight, thanks, I only see his public stuff which is generally pretty good, but I assumed he would put more quantitative/exclusive analysis in his subscriber stuff
Yeah, wow, not what I expected at all. I don't know if I buy what he's saying? It seems to rely on things happening that I have my doubts about, like Ginni Rometty departing.
More than anything I've read recently, this article makes me bullish on Microsoft, who are successfully executing the playbook IBM ran to right the ship in the '90s:
> The problem for IBM is that they are not building solutions for clueless IT departments bewildered by a dizzying array of open technologies: instead they are building on top of three cloud providers, one of which (Microsoft) is specializing in precisely the sort of hybrid solutions that IBM is targeting. The difference is that because Microsoft has actually spent the money on infrastructure their ability to extract money from the value chain is correspondingly higher; IBM has to pay rent
The bearish argument to your bull is that it's fine that Microsoft dominates the hybrid cloud, but where is the growth after that? New companies are entirely serverless with no need for a hybrid solution. So where is the growth after they've converted all their existing Microsoft customers to Azure?
I find the directions the cloud space is evolving in completely fascinating. I suspect we're going to see some major shifts over the next decade.
Primarily: Azure is already super successful and I see it taking a solid #1 over the next few years. The main reason is that Microsoft is doing what they do best: offering a product in every category. Want Kubernetes? They've got it. Need issue management? Azure DevOps. Chat? Teams. Email and Office? 365. Once you're in that ecosystem you're just going to bleed into all of their products before suddenly you're a Microsoft shop.
Moreover, Microsoft has done an extraordinary job at revitalizing the user experience of many of their products. Azure DevOps and Teams has its issues, but they're miles ahead of the products they're replacing, and truly only maybe 10% behind the 'best in class' of each category.
Compare this positioning to Amazon: AWS still only competes solidly in the cloud space, without extending out into the rest of "development team support" that Azure/Microsoft does. They were also ridiculously slow to accept that Kubernetes won the container war, and even today EKS is a comparatively garbage product when lined up against AKS/GKE (no managed worker nodes, $0.20/hour for the control plane, reliance on eksctl to do anything easily which isn't even an official AWS product, etc).
Amazon will be fine, but I worry about Google Cloud's future. They're industry leaders when it comes to new technology, no one else is even close, but that doesn't really matter when its open source and Microsoft is so quick to implement everything they make directly into Azure. Why isn't Google Cloud more successful? I think its enterprise support and respect, which isn't something you can just buy especially with Google's lackluster history in that domain.
The really scary thing about Google Cloud to me is that Google has a very long history of axing projects. I'm just trying to imagine what it would be like to be totally tied into GC and to receive an email along the lines of 'Dear customer, to serve you better we have decided to retire GC by the end of this year, as per your contract you have three months to move'.
Google takes G-Suite/Enterprise focused products very seriously. I can't think of many major examples of them axing one of their enterprise products (or really any, though I'm sure there's one or two fringe examples). In other words, I think they have an issue of reputation here, but that reputation is mostly unfounded because it applies to a different class of products.
I also think GCloud is very unique in that all of the capex that Google spends on it comes back to benefit Google. All of their product offerings need servers and world class networks behind them. So even if GCloud remains a solid third/fourth place in the cloud race, I doubt we'll see them go anywhere just because Google itself is probably using it internally to some degree.
I mostly worry about the scope of investment from management. I want to see way more products out of them. More managed databases, a better managed NoSQL database than Firestore, lower prices, more enterprise focused things like automated security auditing, better support. We're just not getting it; they're so focused on shiny trivialities like Firebase and Stackdriver while missing the basics.
How unfortunate when your audience crosses your carefully laid lines and is active both as a business user and as a private individual. Screw them on the one and expect mercy on the other and we'll see how that multiple reputations theory works out in practice.
"The best thing going for this strategy is its pragmatism: IBM gave up its potential to compete in the public cloud a decade ago, faked it for the last five years, and now is finally admitting its best option is to build on top of everyone else’s clouds."
Wow, that revenue graph is pretty shocking and enlightening to me. I know IBM is a huge behemoth but never knew it had much higher revenue than Microsoft for most of 2000s.
That's the problem with public perception. Everybody talks about the current superstars and thinks the the older companies are declining. If you have read news for the last year it would be hard to imagine that Microsoft just overtook Amazon in terms of valuation.
Microsoft took over Amazon in terms of Market Cap, but not valuation.
MSFT
Market Cap $839.7
Less: Net Cash (59.7)
Total Enterprise Value $780.0
AMZN
Market Cap $834.5
Less: Net Cash (5.6)
Total Enterprise Value $828.9
The discrepancy is all the more evident in terms of multiple (or the "exchange rate" for investing in these companies). MSFT trades at 21.0x 2019E P/E and 12.5 TEV / 2019E EBITDA while AMZN trades at 60.8x 2019E P/E and 19.5x TEV / 2019E EBITDA multiples.
The perception issue is less about the current superstars vs the older companies but rather about the highly visible consumer products vs B2B offerings. Everyone things IBM is gone because it stopped producing products for end users - but the majority of money and development effort is spent on software that's purely for business internal use.
It seems to be a common thing for giant companies to extract massive profits even while the world has shifted underneath them— look at blackberry post-iPhone. It took a few years to collapse.
The Blackberry story is an interesting one. They kept growing in the aftermath of the iPhone through corporate sales because enterprise preferred the tight integration between BES and Active Directory, and they liked that BB phones could have nearly the same draconian IT lock-down policy as office computers.
Even after growth stalled, they had a ton of cash to keep their operations afloat. But they burned up a lot of it in manufacturing the Playbook tablet, which was a big flop.
Interestingly, Blackberry Messenger had 75 million subscribers in 2011, and had a similar cachet of exclusivity as early FB, when signups were limited to those with .edu addresses. CEO Jim Balsillie pushed for BBM to become a cross-platform software product, but his co-CEO wanted to focus on devices and security, and not build software for other platforms. We all know how that ended.
And yet no one really knows what they do. How many regular people have had their life improved in any way by IBM? Compare to Apple's strong stance on user privacy and easy-to-use technology, Google's wealth of free high quality services like Maps, Gmail, Youtube, and good affordable notebooks, and streaming companies like Netflix providing high-quality media content for much lower prices than cable, with little to no ads. IBM can't go bankrupt soon enough. They're a waste of engineering talent and have contributed nothing to the world.
> They're a waste of engineering talent and have contributed nothing to the world.
That's arguably true now...but to claim that IBM hasn't contributed anything to the world is just hyperbole. I mean, they invented the relational database and hard drive, and practically every modern computing concept is just a reimplementation of what existed on a mainframe.
What your man in the street thinks about who has improved/affected their life and who actually has improved/affected their life are in most cases disjoint sets. Ask the man in the street who the most important computer innovators are and they'll say Steve Jobs and Bill Gates (who are/were business-men before anything else), you won't hear them mention John McCarthy or Dennis Ritchie (or likely even Alan Turing).
Apple, Google, and Netflix are consumer-facing entities- most people in the western world are customers of at least one of these companies. IBM does not sell to consumers, so you generally don't hear about "life-improving" things from them, since they don't really need to advertise or compete that way.
Exactly. They're not consumer facing. But I do think that selling off ThinkPads hurts their image as their logo is no longer on devices. Last time I saw an IBM logo on a device was at T.J.Maxx on the cash register that looked very dated.
Interesting read; I wonder if VMWare is subtly in similar position - transitioning their company from providing software which ran on bare metal and essentially managed the datacentre; to trying and not succeeding at their own cloud offering; to now attempting to provide layer on top of somebody else's cloud - currently a successful strategy, but perhaps one that'll see them squeezed out due to costs long term?
Yes that is pretty accurate. The difference is that Vmware is further along in revenue, and has already leveraged their position in compute to diversify into adjacent markets (storage and networking). Red Hat has failed to do that.
And of course Vmware has already been integrated into EMC years ago, quite successfully. Although you could argue the ongoing Dell/EMC saga could jeopardize that... Depending on how well you think it’s going.
However Pivotal CloudFoundry is not their best-seller... I doubt it’s even in the top 5. Vmware’s empire is built on Vsphere/esx (their core business) and, more recently Vsan (storage) and NSX (networking). The former is the largest. The latter two are the fastest growing.
Container platforms are a relatively new category, so sales are not huge. In that category Cloudfoundry might be slightly ahead, but I think Openshift has more momentum because their early bet on Docker then Kubernetes is paying off. Pivotal has PKS but it feels half-hearted.
Pivotal is a separate entity and business model from VMWare, I wouldn't compare them directly. They are both owned by EMC/DELL.
I'm under the impression that cloud foundry is specifically targeting mega corporation like fortune 100 banks. In that context, it only takes a few sales to capture the market and a lot of money.
Another factor that is going to pinch IBM here: they're getting hit from the other side with professional services firms (McKinsey, Big Four, Booz Allen, etc.) now jumping into cloud implementation work. Also smaller but legitimate technical firms like Pivotal who capitalized on IBM dragging their feet in this marketplace.
So, if IBM's strength was "we can do services from anywhere," that gets busted by the rest of these players who can now compete at the same logistical level. And if they aren't hiring the best engineers, they're going to get slammed by Microsoft who will pay top dollar for them.
Doubly so because their investment into RHEL legitimizes the technology for anyone still so conservative as to be unsure. Well if _IBM_ says it's legitimate, it must be so. But really once that barrier to entry is gone (as you said) anyone with enough credibility can sell the tech because it's open source. Arguably unsure people may want to buy Red Hat for support but they'll quickly find it's available elsewhere cheaper and possibly not needed at all.
This makes the openSUSE Leap -> SUSE Linux Enterprise supported upgrade strategy interesting, because if someone started out with openSUSE Leap 15.0, they can trivially attach a SUSE Linux Enterprise 15 subscription to it and convert the system, and remain fully supported by SUSE in the process.
So with that available, if SUSE isn't completely stupid on how to go to market, they'll actually figure out a way to capitalize on that.
Using SUSE in 2018? I don't want to be unfair to SUSE but I would probably put that in the abandoned OS along Solaris, HP and a few other I forgot the name.
They can't kill CentOS. It was an independent project for a long time, and it can go back to being independent. Worst thing they can do is delay the next release and force them to change their name.
Not necessarily sure it'll work out for IBM, but the article brings up an interesting parallel.
In its previous comeback, part of its strategy was positioning itself one layer above the commodity layer. (The article says, "The actual technologies underlying the Internet were open and commoditized, which meant IBM could form a point of integration and extract profits, which is exactly what happened.")
Now with open cloud technologies aiming to make cloud providers (Amazon, MS, Google) into commodities, this new effort also shares that strategy.
Of course the other ingredients are that it must be a value-add that customers actually need and nobody else offers a better version of. But buying Redhat seems like a reasonable stab at achieving that. Overall, their strategy seems to make sense at a very high level.
As someone who has had to deal with two cloud-phobic industries now (auto and healthcare). I have to say the hope of the hybrid cloud being a differentiator, for RedHat/IBM, is misplaced.
Getting these players into the cloud is always hard, but the hybrid cloud doesn’t really soften the ground internally at most of these compliance heavy organizations. They are all aware of how awesome and wonderful the cloud is, it’s always politics that get in the way. Once an organization goes through the incredibly difficult process of allowing cloud projects (which involves lawyers, rewriting policies, compliance officers, etc), the organization usually puts all projects going forward onto the cloud and views the old stuff as legacy. The idea of trying to bridge the two is rarely considered worth it.
Of course there are exceptions. However where I see hybrid clouds do well is in organizations that are taking the task of porting everything to the cloud very seriously, but they know the project will take a while, so it is worth everybody working off of the same infrastructure playbook until they can shove everything in the cloud.
The days are dark for the folks that missed the cloud bandwagon over a decade ago.
I guess what IBM needs to do is to go talk to its customers.
Server-less is an interesting development. To avoid lock in though we would need a library API layer. Much like what Kubernetes provides at the application layer.
At this point, is IBM anything more than an accenture/infosys/cognizant style consulting firm with patents? I'm having a hard time understanding why they're supposedly different compared to any of the big players in that space.
IBM today is Unisys of 1990, it's 2017 did you know that Unisys still sells Mainframes?
IBM has become completely dependent on consulting, and they did that because their hardware business has been shrinking. They are only selling hardward to a captive audience. The only companies buying Mainframes or AS400's are existing customers, but those customers are also converting portions of their business to other technologies, but not on IBM hardware, or the IBM cloud. It appears that IBM is still a technology company but it's an illusion.
IBM is dying and there is nothing that anyone can do about it, they won't fold, but in 20 years IBM will be as relevant as Unisys is today.
I work on IBM clouds all day and what they have been doing for the last several year(s) is embracing Linux, much the same way that Microsoft has been embracing Linux with containers, WSL, github, and Azure.
The first thing I thought when I read this news was - good maybe they’ll take their 360,000 employees, make them all use Linux, and as a consequence standardise a developer laptop to compete with MacBook.
The second thing I thought was - if there is 360,000 people with Linux laptops in your enterprise, what sort of creative uses could you come up with for container orchestration.
The third thing I thought was - I wonder how much they pay in Microsoft licensing per year.
Not talking about starting a laptop business - talking about a major corporation with 360K people that are all-in on Linux. Someone needs to fulfill that hardware requirement - it doesn't have to be IBM. What does that sort of requirement do to the IT ecosystem.
Off topic : Ben's articles on Stratechery were pretty much only becoming about one or two things only i.e Bundling and Aggregation. I am happy that from the Instagram's CEO article to this one all of the posts did not disappoint me at all and are not a unhealthy repeat of the same themes and or theories.
If you force-divested IBM into constituent parts which one(s) would retain the IBM branding and why?
PowerPC was the last time I was aware of IBM making significant change in the VLSI market. Before that, it had been definitional in a number of things, not networking (!) but certainly high density storage.
"This is where the Red Hat acquisition comes in: while IBM will certainly be happy to have the company’s cash-generating RHEL subscription business, the real prize is Openshift, a software suite for building and managing Kubernetes containers."
Small problem with that. Most of the progressive enterprises are trying to move AWAY from containers to fully serverless architectures (Lambda, Google functions, Azure, etc.)
Kubernetes is still hot but the momentum is definitely with Lambda and its ilk.
I'm a consultant who works with enterprises of all sizes (all sizes that could still be considered "enterprise" sized at least) and I've only encountered one that has used anything serverless, and then only in one production application. And I touch basically every IT application/environment across the entire business.
Now I don't want to say my experience is perfectly indicative of what big companies are doing in general, but I certainly haven't noticed this trend. Even containers are just starting to catch on in the enterprise market, and a lot of enterprise software doesn't currently support containerization.
>>Most of the progressive enterprises are trying to move AWAY from containers to fully serverless architectures (Lambda, Google functions, Azure, etc.)
Serverless runs on?
Also Lambda is not a replacement for your webservices. Lambda is very expensive for non-trivial scaling requirements. And yeah bulk of the backends in the industry aren't always web based products.
Redhat isn't going to save IBM. They sold because they had an unsustainable business, and Openshift is still a risky play to run a stable business on. The container space is starting to get awfully crowded, and cloud providers can provide better tailored experiences and skip containers altogether with serverless.
As a long time taker of Redhat certs, I hope they pull through and still look something like the Redhat I have loved and known.