HP To Shut Down Its OpenStack Based Public Cloud (fortune.com)

An anonymous reader writes: Hewlett-Packard, which has been backing off on ambitious public cloud plans for a year, is now calling it quits, sunsetting HP Helion Public cloud in January 2016. in a buzzword-laden blog post, the company says its building out support for interoperability with Amazon and Microsoft public cloud offerings to provide options for customers who require such functionality. "HP’s decision is the latest milestone in what has been a slow fade for the company’s public cloud ambitions. It has become increasingly clear that there are three, maybe four companies that can support (at scale) the massive shared computing, networking, and storage infrastructure necessary for a public cloud. ... HP will continue pushing its private and hybrid cloud."

Comments

More accurate ...

[gstoddart]

HP's decision is the latest milestone in what has been a slow fade for the company

HP has been in decline for years.

Quality is down. Innovation is down. A series of seemingly incompetent CEOs. A couple of bad purchases. Some stupid decisions. Some utterly failed products.

Like so many large companies, now they mostly just lurch from one thing to another hoping sooner or later one of them sticks. One gets the distinct impression nobody really has a clue of what they're doing, and even less of a clue about what to do about it.

Welcome to the modern world of tech, where you buy everything in sight, fuck it up, have a bunch of bad management, and then eventually implode as you realize nobody in your organization measures up to the people who got you there.

One wonders how many good companies have been swallowed up and ruined in trying to make huge companies more profitable, only to find out the huge companies have no idea what they're doing.

Over and over again, we see big corporations who really just keep changing CEOs, and utterly failing to understand just how badly they're all screwing up the company.

Re:More accurate ...

[gstoddart]

Well, "which HP?" is a good question.

As Hewlett-Packard (HPQ - Get Report) prepares to break apart its enterprise unit from its division that produces personal computers and printers, the tech institution said Wednesday that it will sell cyber security unit Tipping Point to Tokyo-based Trend Micro International for $300 million.

Splitting up and selling off are not exactly indicators everything is going swimmingly.

As I said, companies grow, companies buy, companies fuck up what they buy, and then realize they no longer have much of an idea what they are anymore.

I think the M+A craze in tech for the last 20 years has been a lot of short term profit seeking, but is overall really bad in the long run. It maximizes executive compensation, but it doesn't actually achieve the outcomes they claim it does.

Re:More accurate ...

[RabidReindeer]

HP. Downsizing its way to Greatness.

Re:More accurate ...

[Anonymous Coward]

HP is an ink company.

Re:More accurate ...

[Anonymous Coward]

Here is an anectdote, but a good example.

Small-ish company, startup culture. Not the kind of startup culture where there is a ping pong table, the kind were we feel we're the underdog in taking away big corporate accounts for a niche product where the big players have basically shut down development for nearly a decade. Needless to say, we were estatic that we could walk in and basically offer a redo product, and that alone gave us advantages.

Eventually we grew the company to a 40 million per year revenue generating company. Along comes Cisco, who wants to "buy their way into the server room" in the operations software side of things. Now the product is a tiny hard-to-find offering, and it basically still exists; but as an improved way to manage Cisco offered hardware. That's because Cisco doesn't know how to sell software, and so we were pushed to be a ticket rider to their hardware sales, or risk losing any means of selling what was a growing product.

After a few years, the cuts came, and now there's basically a skeleton crew to maintain what was once the flagship offering. The product became so obscure that it even dropped off Gartner's map in comparison to other products in the industry. I'm sure it's making money, but at many sites I imagine it is shelf ware. You know, the stuff that was along for the ride, and you'll install it one day to try to figure out what it is. Then you'll decide it's pretty cool, but you really don't have enough time to learn it (and don't want to take the risk of depending on what you don't understand).

Re:More accurate ...

[Junta]

Either way it's a shell game for the investors. The problem of being a publicly traded company in this position is that they have to do things to appease shareholders that are actually pretty dumb.

Case in point here, splitting their enterprise x86 stuff from their PC stuff is mind numbingly stupid as hell. IBM did it and it seriously screwed up their x86 server stuff. You can do a bit better than break even on consumer space, but in the process you have massive negotiating power with component vendors. Intel wants in on a tablet play, sure, but need price breaks on Xeons for the opportunity. Hynix wants to sell modules into your laptops? Sure, but need a price break on RDIMMs.

Splitting it up may help to settle disputes about who is doing what margin-wise versus revenue wise versus total profit, but it's pretty good way to cripple your supplier negotiating power, which is really what matters here.

Re:More accurate ...

[gstoddart]

Slashdot is so weird. Big companies are, for the most part, evil [unless they're run by young, hip, smart guys who promise that they won't be evil].

No, we're pretty consistent ... we're just as convinced the companies are evil when they're run by young, hipster douchebags, and we don't believe them when they promise to not be evil.

Long established companies have just managed to grow into lumbering beasts with no clue of what they're doing, but who have repeatedly established their evil. The ones ran by the young hipster douchebags are just more focused on a narrower field of business, and can pretend to not be evil hipster douchebags yet.

Lots of us think it's just a matter of how the evil manifests itself.

The small lean but evil corporation is evil in a different way than the huge organization which isn't sure how to remain relevant.

Make no mistake, all companies converge on evil.

Re:More accurate ...

[Solandri]

Like so many large companies, now they mostly just lurch from one thing to another hoping sooner or later one of them sticks. One gets the distinct impression nobody really has a clue of what they're doing, and even less of a clue about what to do about it.

Nobody at HP really has a clue of what they're doing. The people who had a clue were either let go or were spun off as Agilent Technologies, which inherited HP's instruments and test equipment business. That's where HP's soul went. The company called HP today just kept the HP name.

Re:More accurate ...

[Icyfire0573]

I think what you mean to say is that they need to make Tough Choices :p

HP = Printers

[xxxJonBoyxxx]

As a brand, HP means nothing more to people anymore than "some company that makes printers."

For it to thrive for the next 20 years, what HP should do is:
1) Get some consumer/business cred (and market share) back by selling "the longest lasting print cartridges"
2) Buy up the 3d printer market and develop a brand that builds on "InkJet" and "LaserJet" like "3DJet"

Re:HP = Printers

[Marginal Coward]

As a brand, HP means nothing more to people anymore than "some company that makes printers."

Speaking of HP as a brand, I'm sure many here remember when HP meant "the best test equipment money could buy" (except for oscilloscopes, of course.) That sort of reputation for a brand is rare indeed. However, after going into computers, making printers, merging with Compaq, etc., they spun off that business and renamed it "Agilent." Later, Agilent became "Keysight" after they split that into a test equipment business and a medical equipment business, the latter of which retained the Keysight name. But whatever they're calling their test equipment business now, I think they've lost significant brand value, regardless of the value of the products that Keysight actually makes.

Evidently, they thought the "HP" brand in computers - or more likely printers - was so strong that they would retain that for those products at the cost of losing the top brand in the test equipment business - which, of course, is the business that Messrs. Hewlett and Packard originally created.

Today, the PC business is a commodity business that nobody wants to be in anymore for that very reason. I suspect the printer business eventually will be a commodity business, if it isn't already. Of course, brands don't much matter in commodity businesses.

Sayonara, "HP," my old pal - you were a good brand while you lasted...

Re:HP = Printers

[Marginal Coward]

But what about buying from:

4) Shell, even though it costs more, because Shell is the brand that sells the very best gasoline.

Re:HP = Printers

[Marginal Coward]

You're right that most people consider a variety of factors when buying a commodity product such as gasoline. Brand may even play into that a little - after all, the people who sell gasoline certainly advertise to maintain their brands. But the fact that brand didn't appear in your list illustrates the fact that brand is pretty small among those factors. Commodity business have to differentiate themselves by practical factors such as the ones you list as well as price (for most people.)

In my own case of brand loyalty, I prefer Coke, but I don't regard it as a hardship when I end up with Pepsi instead. And whenever I ask for "Sprite" at a restaurant and they ask if it's OK to instead give me whatever similar product they happen to have on tap, I always say "yes."

Re:HP = Printers

[AuMatar]

I worked at HP when they transitioned to cheap printers. At one meeting a manager stood on top of an HP printer while it was printing, and it continued to work... then used this to claim we're making too good a product and should be making cheap consumer shit.

They used HP hardware to build their cloud

[Anonymous Coward]

They used HP hardware to build their own cloud because it was cheaper.

Then they found out they needed to buy separate ILOM cards so they could manage their hardware. So much for cheaper.

They spent hundreds of man-hours installing the ILOM cards.

Then they found out they needed to by licenses for the ILOM software. Oh crap, now it's more expensive.

Then the servers went operational and they found out their brand-fucking-new HP servers are actually SLOWER than the years-old Sun Microsystems-built x86 servers they're supposed to replace - because HP used cheap parts like motherboards with no memory bandwidth and complete piece-of-shit commodity disk controllers that shit the bed when trying to move more than 50MB/sec. Oh yeah, the NEW HP servers only have two gigE ports - whereas the OLD Sun servers had FOUR. And if they had used new Oracle-built servers from the old Sun production line, they would have gotten four 10GB network ports on FASTER servers - and had remote management hardware with fully licensed software for LESS money. (And Oracle ain't exactly cheap, so when the final solution from HP is shittier and more expensive than the one from Oracle...)

Then the damn HP hardware started failing right and left. And the cheap-ass fiber-channel HBAs don't work with a damn.

Can you tell I've dealt with HP before?

So it's not surprising the HP can't deal with HP either.

Re:They used HP hardware to build their cloud

[Anonymous Coward]

I used to work for HP and you are not far off. The cost center accounting meant that group A had to pay Group B for the product. Mind you it was only on paper but still there was a markup at each step.

Ill also point out that there data-center services are crap. I pointed out that I could automate a lot of simple tasks that would reduce down time and provide the customer better service. The Director asked if it would reduce ticket counts as each of these items generated tickets. I said "Yes, it should cut the tickets in half, reduce maintenance times, increase up times, and make for a much more manageable service."

I was directly ordered to NOT do it, Told to never bring it up again, and that if I ever let the customer know I would be fired. Seems they are paid by the ticket. So anything that would reduce the number of tickets is a bad thing, even when it is in the customers best interest.

I was also told to tweak monitoring and make it more sensitive. That way it would create more tickets and more cash flow.

Thats what you get for outsourcing to a data-center, they are not looking to provide your company with great services. They are looking to make more money from you.

MBAs + H1Bs = HP

[zerofoo]

HP will be the poster child for what happens when MBAs and H1Bs take over a company.

Ultimately, tech companies need to be run by tech visionaries. Car companies need to be run by car guys/gals. Financial companies need to be run by sharks.

You can't simply crank out an MBA and put that person in charge of a bunch of cheap programmers and expect innovation. Creativity and passion can not be taught.

I miss the old HP, run by passionate engineers, that built the worlds best calculators, printers, and oscilloscopes.

Re:MBAs + H1Bs = HP

[ErichTheRed]

"You can't simply crank out an MBA and put that person in charge of a bunch of cheap programmers and expect innovation."

This. It boggles my mind that no one has stood up and said "this is stupid" over the last decades. Having worked in big corporations almost exclusively, I've seen a lot of this:
- Management by spreadsheet, exclusively. (If you can't measure it you can't manage it.)
- MBAs being put in charge of divisions they know nothing about. (A good manager can manage anything.)
- CxOs listening to McKinsey, BCG, Bain, Accenture, Gartner, etc. exclusively. (They're the most expensive so they give the best advice.)

The last one is particularly funny -- I've experienced a lot of white-shoe consulting firm employees coming in and telling veteran execs 20 years their senior what to do. All of these firms' business models are to take newly minted MBAs, fly them to client sites and dazzle executives with sales PowerPoints. As you can imagine, the problem continues until the consulting firm can't extract any more fees for those former students.

Re:MBAs + H1Bs = HP

[gstoddart]

This. It boggles my mind that no one has stood up and said "this is stupid" over the last decades.

Well, they're in a feedback loop now .. the MBAs are the ones saying you need to bring in MBAs.

Much like how CEO pay climbed to stupid levels because CEOs and their cronies sit on multiple boards and decide how much to pay CEOs, when the sit on multiple boards and decide how much to pay CEOs.

The people making the bad decisions are hiring people just like them to help make the decisions, because thay mostly want like-minded people. Their compensation is separated from their performance, and they only look at the next quarter so they can take their bonuses and run like hell.

Corporations are almost incapable of long-term planning, because they really only give a damn about the next 3-4 months, and ensuring they maximize their own compensation.

My personal opinion, like you having seem it for the last bunch of years, is that C-level management and MBAs are pretty much uniformly fucking up corporations by doing the same stupid stuff over and over again. But by the time anybody realizes these clowns have been promoted, or received their golden parachute.

Other than CEOs, who really believes CEOs are worth anywhere near what they get paid? Oh, of course, the people who aspire to be CEOs so they too can get overpaid.

It's the emperor's new clothes. It's lies and stupidity piled on top of one another.

Smart

[willworkforbeer]

At this late stage, trying to catch up to AWS or Azure would be a death march. They should try something with better odds, say, invading Russia during the winter.

It's not HP...

[Anonymous Coward]

I work for a fairly large service provider, we've been trying to be successful at deploying a full public cloud based on OpenStack for a long time now (years). Today, OpenStack is put together by a fairly large amount of private cloud folks. It's a very nice offering if you want to run a private cloud -- even a very large one, but running a public cloud is different. You have to think about multi-tenancy -- scaling in the number of different segregated users using the cloud not just the number of nodes in the system, you have to think about no down time deployments and migrations, etc. OpenStack is just not focused on this -- and there have been a lot of design decisions that make running a public cloud a pain. We've tried multiple times to submit patches that would enable public -- but there is a lot of bureaucracy and control by private cloud companies that most of those patches fail to make master. Thanks to DefCore [https://github.com/openstack/defcore/blob/master/doc/source/process/CoreDefinition.rst] if we introduce our own patches that fork core in any way -- even for our own deployments -- we risk losing the OpenStack trade mark. So it's very difficult, we haven't given up on OpenStack public cloud..yet...but it's frustrating.

Re:Smart move

[BenJeremy]

If they want to dive into a market, they have to be in it for the long run, just as Microsoft is prepared to do. Amazon is leveraging a lot of their existing resources, which helps them; but Microsoft is probably bleeding pretty badly in the short term, and that's OK.

HP keeps doing dumb things, jumping in head first into a market and mismanaging that effort to a bitter, short end. Part of the problem within HP is that while they can easily manufacture that cheap disposable whitebox hardware to set up the required data centers, the company is run like Lord Of The Flies - with groups charging FULL PRICE for products and services between other HP groups. There isn't even a chance to negotiate terms, and some groups are faced with outsourcing things as simple as a web site for internal support, because to do it in house would cost them 10 times what Azure or Amazon will. You can bet that those server blades to run the data centers (which would also be owned by yet another chunk of HP) will cost exactly full retail, instead of leveraging their resources and providing them at cost (which helps everybody, since the increased production drives down costs and increases profits on outside business). Groups will simply not cooperate within HP, and that kills the bottom line.

Who would use HP anyway

[EmperorOfCanada]

As a tecnology person I have only ever seen HP as a trio of companies. One sells crappy laptops, one sells printers with exploitive prices for ink and toner, and the other has salesmen in cheap but still too slick suits trying to sell services that only make sense to techno-illiterate CEOs and board members. Not once in my life have I heard some respectable IT person put HP on any list of server technologies that they were even looking into, let alone buying.

The same when I used to see any business that had all IBM desktops. I would just laugh and say, "I wonder what the total cost per machine that is going to come to?" You could be certain that those machines were just part of a giant upsell as soon as some IBM saleman got his foot in the door during a golf game with some executives in the sucker company.

So if anyone is put out by this decision to close this technology it just certifies them as a fool.

Unravaling Carly's legacy

[plopez]

Dumping the Compaq crap. Now they need to dump off EDS as services is dying as well (for many companies, not just HP). They are slowly and quietly putting a knife to services as we speak.