HPE campaigns against 'cloud first' push in UK public sector

Hewlett Packard Enterprise has posted a "UK Public Sector Manifesto" with nine themes, alongside a campaign hyping the value of hybrid cloud. The bugbear for HPE is that UK government introduced a "cloud first" policy in 2013. The current version was revised in 2017 but it mandates that central government, when buying new IT …

[Anonymous Coward]

"Since "cloud first" already allows for non-cloud deployments when justified by cost"

So everything is going to be on prem then, if justified by cost alone, only small or extremely bursty deployments can be justified by cost alone to go into the 'cloud'. The 'cloud' isn't cheaper, mid to long term, it just looks like it at first until you actually use it.

"but cloud still offers many possibilities such as auto-scaling and serverless computing that do not exist in the same way on-premises"

Really, doing it perfectly fine on prem.

[Anonymous Coward]

Ref the last point, I think what they were driving at was more around large scale, short term burst capability.

Take for example a product I'm familiar with. 99.9% of the time it'll eat level of resource X. Very occasionally it will require Y which is significantly higher, but also highly variable as to HOW much higher.

In <cloud> we can just scale it up on the few times it needs it till it's churned through the load. On prem we'd need hardware sat idle most of the time to cover those few situations as it's usually time critical.

Of course there's ways around it, but there's definitely cases where it's not worth focussing the dev time on rewriting the tool to handle it differently for the minimal impact it has. Of course it's already a cloud solution anyway and I'm probably explaining it super badly by trying to talk in generalisations.

I guess the Tl;DR is that sometimes it works much better and is more cost-effective than maintaining many racks of mostly idling hardware. It would be like me running a Konigsegg as a daily for 100k annually so as I could take a once yearly trip to the Nurburgring.

[ecofeco]

Exactly. Cloud and on-prem are like having closets at home and public storage for the extra.

Who in their right minds puts ALL of their possessions in public storage as a permanent thing and tries to live like that daily?

Data is not like physical inventory. It is the very blood of every business. Insurance can replace physical inventory that is no longer accessible for various reasons. It cannot replace unique data.

Cloud is a great place for back up or surge demand. It is a very, very stupid thing for primary reliance.

"Here stranger, these are the keys to my house." How is that a good idea except to idiots?

[claimed]

Like renting? A publicly accessible, but occupied house. Landlords with scale (e.g. block of flats). Lots of people rent, they're not idiots.

You were making an ok point but messed it up at the end.

[Anonymous Coward]

No, they are just people that cant afford / able to get a mortgage. If you can get one and have the money, you would be buying a house, its more profitable, if they don't, they are idiots (unless the collapse is coming soon).

[claimed]

I can think of multiple reasons why you might rent even when you can afford a house. That's beside the point.

Don't call people idiots because you don't understand their motives.

Chestertons fence.

[elsergiovolador]

Conflict of interest

If you have infrastructure underpinned by a company being accused of not paying fair share of tax, then you are in the pickle.

If you push to make them pay, then you risk they will pull the plug.

[Pascal Monett]

"central government [..] must consider [..] a public cloud [solution] – before any other option"

Simple answer :

This option has been considered and, after much research (and an exploratory team sent to the Bahamas to witness cloud deployment on-site), the conclusion that imposes itself is that Cloud is not secure, not fit to host citizen's private data, not reliable enough, prone to lock-in and more expensive than it seems.

Therefor, for this contract, an on-prem solution will be preferred.

And there you go, problem solved.

[Mike 137]

"cloud first" already allows for non-cloud deployments when justified by cost

Maybe we should rephrase that as ' "cloud first" already allows for non-cloud deployments when justified by initial cost'

Once they have you by the long and curly contract, the price can be hiked at will (it really does happen). And it can be darned difficult (and expensive) to migrate to an alternative provider, particularly if you rent SaaS - due to compatibility problems.

[Anonymous Coward]

The sad calls I take from HPE that put a smile on my face

After years of rapidly increasing HPE support costs we went cloud only, Not cloud first but cloud only. We decommissioned rows of racks of HPE kit in multiple datacentres and never looked back. HPE sales call me every six months trying to extract some server sales "hopeium" for their quarterly targets. They always assume that we have some legacy hardware, some last bastion of on prem kit, some unsolvable legacy app pain point. We don't. Explaining that we have done it, cloud only, not hybrid and that hardware uptime is all now someone else's problem is joy I am yet to tire of. Thousands of employees working either from home or the hundreds of offices we support. It isn't perfect but it is better.

[dgeb]

Re: The sad calls I take from HPE that put a smile on my face

I’ve never really understood the point of support contracts on commodity servers. It’s both cheaper and quicker to just buy a [few] spares. (It’s a slightly different matter for one-off big expensive things, where they are both complex and n+1 is a heavy cost burden, of course).

That’s especially true if moving to cloud is on the cards, as that already requires you to have [re]architected stuff to accommodate instances failing and to avoid strict dependencies on any single bit of hardware.

[BlueInfra]

Re: The sad calls I take from HPE that put a smile on my face

Long ago (pre-virtualization) there was value in rapid repair of on-prem servers. I had a couple of important HP servers that I bought a "6 hour call to repair" contract on. In theory HP was guaranteeing that within 6 hours that thing was going to be fixed.

[dgeb]

Re: The sad calls I take from HPE that put a smile on my face

If those important servers are also unique in hardware terms then that makes sense (this is what I was getting at with n+1 being burdensome above). If they’re otherwise the same as a chunk of the rest of the fleet, I’d rather have a spare in inventory - which can be swapped in to service in under an hour just by either swapping disks or configuring the HBA to match. If the thing that broke is under even basic warranty coverage, it can be sent off for repair/replacement to replenish the spares inventory.

We have a hardware support contract on our tape library - that has value to me because it is a fraction of the cost of buying a whole second one as a spare.

[RichardBarrell]

Re: The sad calls I take from HPE that put a smile on my face

I assume that SANs are also in the set of things you want a hardware support contract on?

[dgeb]

Re: The sad calls I take from HPE that put a smile on my face

I think with vendor SANs it's more the software that needs support - the hardware (whether a vendor designed SAN or an in-house storage platform) should have enough redundancy and modularity that spares work well there too in a deployment of any significant size (in that a spare disk shelf, controller, power supply, a few cables, and a handful of drives are all you need).

If a vendor is involved, of course they probably will insist on having hardware support to get the software support, so you may be stuck with it.

(This is also sort of an argument for having a few storage platforms instead of a single massive SAN - the spares spread across more instances - but you also mitigate the eggs-in-one-basket problem, and are less likely to be running into the edge cases than when near the envelope of the platform's capability.)

[RichardBarrell]

Re: The sad calls I take from HPE that put a smile on my face

Thank you.

[Giles C]

Re: The sad calls I take from HPE that put a smile on my face

A previous employee used to do that with wireless APs, the controller was on maintenance as was one of each type of ap. sitting in the storeroom was a box of about 10 APs ready to be deployed if one should fail.

I think they were a couple of hundred each and the maintenance on one was about £40 per year. As the deployment had around 150-200 it was a lot cheaper.

[cantankerous swineherd]

this on the day when all the border farces biometric scanners went down.

[NeilPost]

Which was caused by ……. ??

[Peter D]

Economic Meltdown

My only worry about the cloud is the possibility of a devastating infrastructure attack by a state actor to destroy the processing capability of the likes of Azure and AWS. A comprehensive attack could create an economic meltdown larger than that caused by a few nukes.

[RegGuy1]

UK GOV: MUST do public cloud only?

The current version was revised in 2017 but it mandates that central government, when buying new IT services, must consider a cloud solution – and specifically a public cloud, rather than "a community, hybrid or private deployment model" – before any other option. (my bold)

This is just cost saving. Are you, government, going to mandate where in the cloud the kit is hosted -- must it be in the EEA? Are you going to mandate who can manage it -- no Indian or other global access to sensitive data/bases? Are you going to insist on DOS protection?

A stupid mandate by idiots who know nothing. Surely, Shirley, you would want at least your databases on prem and have some form of access control from your cloud frontend. The number of government contracts I've seen where you must only have UK support staff access the data (data protection) and which has pushed up the costs are now no longer required? To save money? Because we've left the EU so you can claim it was them wot stopped us?

What happens when the cloud inevitably goes down and even UK staff have no private (internal) access? Hybrid was designed to meet enterprise and public sector concerns. Jesus.

Lead by donkeys or what?

[VicMortimer]

Absolute insanity.

Governments should have a "Cloud Never" policy. Under no circumstances should any government put anything on any server out of their direct control and ownership. It's not like they're a small business and aren't big enough to have their own IT department.

[Anonymous Coward]

Back in the late '80s, HP and IBM were companies to be reconned with, shifting bullter-proof hardware and vertical-strategy operating systems.

Now they're both has-beens desperately trying to find a place in a new world that sees their models as "obsolete"...

[MTimC]

Cloud vs On Prem

I looked at cloud economics in some detail for organisations with annual IT spends of $!M to $7Bn. The best summary of how the value chain hangs together is in Nick Carr's "The Big Switch": on the supply side, capex is a function of maximum demand, operating profit/efficiency is a function of average utilisation. So it pays to pool demand, much like traditional electricity generation based around large power plants.

However, cloud first entails some very significant changes to how IT is built and run, largely as the on-prem model relies on hardware to provide various availability/recovery processes, whereas these are delivered by the application layer in the cloud, you have to manage the demand and capacity used as you didn't buy orders of magnitude too much to start with (and I did measure some instances where, because capacity had to be estimated before demand was understood, such overcapacity had been bought and recovering the costs was being hidden in the budgeting process.) In practice, 'cloud first' implies Continuous Delivery.

Unfortunately, even for traditional IT, govt. procurement isn't always great, and typically creates zero sum 'multi-sourced' contracts, which obviously lead to excess costs.

It's typically not practical to lift and shift existing govt. IT systems as they don't have any tests and assume a particular technical architecture. A good example here cold be HMRC's CHIEF system, which was initially considered for retirement around 2010, but the first suppliers couldn't work out what it did (and they are very competent people). It's currently being superseded by CDS.

Where I looked at on-prem vs cloud security and reliability, the usual situation was that the on-prem characteristics weren't known. When you start digging, you quickly find that the reason that they are not known is to protect the guilty.

[Steve B]

Haven't got the staff any more though.

The Government policies have basically removed most of the intelligence and the best programmers from Local and Central Government employ.

If we wanted something done for our users we did it. It worked, and if we could we sold it to other Councils.

Then the outsourcing came in and after a change of the original management team, who had reorganised themselves into a nice redundancy or early retirement package, the new management team perfectly matched the new psychometric test requirements, whilst still being absolutely useless at everything bar attending meetings and agreeing with the perceived main person in the room!

Instead of the users getting what they wanted, they now got "bespoke, off the shelf" packages, which saved all the high development costs.

The fact they didn't actually meet any of the criteria was glossed over as unimportant, until the "users" started going sick with stress caused by the new systems and it got into the press.

The trouble is that once you have lost the real talent, you are stuck with the dross and can't recover,

Hence all the bills went up and none of the systems satisfied the users, which in our case also lost all off the external customers we had built up over the years, who were paying "real" money into the Council coffers.

The old management team would not touch "consultants", preferring to spend the money on training in house personnel. The new team, being completely incompetent, relied upon external consultants.

We found the first problem was that the "consultants" hired had usually read the book the night before, giving them "expert" status in the eyes of the new psycho management teams and our local relevant knowledge was ignored.

We were always found to be right, usually 6 months into the project, which led to a management team led restructure, rendering certain people surplus to requirements!

The management team performance bonus levels went up afterwards as they could now pass the buck.

When I dealt with Central Government, I found they suffered consultants exactly the same way.

In fact watching the pandemic fiasco and the Government advisors, I see it is exactly the same even now.

[Anonymous Coward]

Good

Unless I’m mistaken, and I might well be, half of the people fronting cloud companies in the U.K. are ex Senior Public Sector.

So I for one welcome a strategy like this being challenged.

Not only is ‘cloud first’ theory extremely questionable, it also point-blank does not work for some situations - and not because of the cost.

[RichardBarrell]

I'm laughing, personally

This is hilarious comeuppance for (a descendant of part of) the company that claimed to have invented "Grid Computing" but then never bothered to productise or monetise it.

[Anonymous Coward]

Re: I'm laughing, personally

Those long winter nights must fly by at your house.

[RichardBarrell]

Re: I'm laughing, personally

Only when t'internet is working and I can look at videos of dogs being hugged.

[Anonymous Coward]

HP DC story

Years ago, I was working for an extremely toxic and non mature company. 9 BUSD turnover per year, 80 k employees.

There was this central DC, serving all employees worldwide. This was an HP shop, Itanium and eyes watering expensive

support costs.

One month before leaving, I deployed Cacti on the 250+ HP(E) physical servers to see what was the load across, say, one month.

Result ? 1 of them was maxed out 24X7, 8 were, say, running workloads. And the rest ? Fuck all, 0, barely syslog would wake up a couple

of times during a one month period.

It turned up, as I was suspecting for years, IT mgmt was so retarded they were randomly allocating CAPEX to every single business new wonderful application that got installed and never pushed to the business. One app = one server. And no-one was using them !

So, yeah, after years of selling millions of Itanic idle servers, I can understand HPE being quite nervous about cloud, where you monitor every consumed hour !

[snipsnip]

STEVE HOLT!