UK government spends another £1B on cloud migration and services The UK government has awarded a contract worth up to £1 billion ($1.3 billion) to get tech services companies to help various bodies and departments make the leap to the cloud. The Crown Commercial Service, a unit of the Cabinet Office, has awarded another chunk of the G-Cloud 14 framework, under which a maximum value of £1 …
So I work in IT for a small local council, several hundred employees serving a couple of hundred thousand residents. Why do I need to push everything to the cloud? There's a bit of a budget problem at the moment as you may have noticed, so cost is absolutely my watchword. Cloud is definitely not cheaper than what we're doing at the moment, so that's out. The main benefits of cloud seem to be connectivity and scalability. Well the number of residents in the area is pretty fixed, we're not going to have a million or so people move in over night, so rapid scalability isn't necessary. All our staff are pretty much within commuting distance from the building, so distributed hosting would gain us nothing, and our infrastructure holds up well to WFH needs. Security is a concern, but as we know, cloud is by no means secure, and presents a much, much larger available attack surface than our set up.
So, why should I be spending extra money on cloud at a time when we really don't have it to spare? Genuine question, I really would love some insight to prove me wrong here.
I see the cloud as a vehicle for certain (erratic) workloads, for which it is very good.
I've seen people move off the mainframe and the true costs expand.
You used to have two people responsible for a few (large machines). You now have many more people responsible for many cheaper machines, and keeping them up to date with fixes etc.
Taking backups and doing disaster recovery was handled by a few people. On the cloud it is much harder, because there are more machines etc.
One customer said they had spare cycles they were not using on their mainframe. The CIO said improve your testing, by testing higher volumes and covering more areas.
The pendulum will swing. "We used to rent time on a machines. They we brought it in house to make it easier and cheaper to mange, now we rent time in the cloud"... guess what will happen next.
Cloud has been good to make people look at their processes and improve them.
It is different to last time in some ways. Cloud has low entry cost - that is a great enabler for new business. The costs can be good but takes a lot of careful design and that is a big upfront investment if you have lots of legacy. But the real reason cloud is pushed is that we are in the middle of a war for the world, not a hot war. There is a move away from democracy to a global collectivism driven by shady characters but executed by our own governments. This globalism (sounds fluffy - it's not) demands centralisation to enable control and that is behind the extreme drive for a few big IT players.
Cloud is someone else's computer. That someone else is very cool, has all the great toys and is popular.
But real answer is probably simple - corruption. Most likely the usual suspects convinced the decision makers, over champagne on a brown envelope coaster, that the cloud is the solution to all the problems.
So billions will be spent and most of the work will be done by a couple of blokes on £60k.
Becsuse it's a buzz word that makes CIO and CFO feel important & gives consultancies & MSPs a way to shaft government for more money.
I worked on planning a DC move, cloud was 300% more expensive than new DC space, new hardware AND the cost of migration. Yet Gartner & CCS STILL couldn't understand why we didn't want to go to the Cloud. And that was BEFORE the latest few years price increases in cloud.
Councils especially run very stable very important workloads that don't change much, so won't gain Anything from server less or the abilities that cloud provides. Add to this, the INSANELY cheap deal CCS negotiated for datacentre space at CCS DCs, got councils cloud is a massive money sink.
>Councils especially run very stable very important workloads that don't change much, so won't gain Anything from server less
But cloud also offers near-instant global location portability.
So if Solihull town council suddenly needed to move operations to Hawaii then AWS/Azure/GCloud etc could accommodate them much more seamlessly than moving a mainframe.
- Seriously at an early SUN (that's how long ago) cloud conference once, Rolls Royce Aerospace were extolling the virtues of their new cloudy system which meant that if they wanted to move their jet engine business from Derby to anywhere else in the world then their IT could move with it almost instantly.
.. which in this context means absolutely zero.
Rule 1 of any of these evaluations is that you first need to define your use cases, that gives you a basis on which to decide.
"The Cloud" was deliberately kept as nebulous (sorry) as possible in order to sell it to people without a clue who could then pretend to be knowledgeable, but at the coal face you still have to have insight in what you're buying, and for quite a few situations remaining on prem proved to be more cost effective or more suitable. There isn't one answer for all, however much especially sales people like to pretend there is.
The DFE are the same, they're pushing for all schools to move to the cloud.
When we costed it, sticking internally with a SAN and two hosts will cost us approx. £5000 a year (over 6 years of server life), not including electricity costs of course (est. £1000 a year)
Hosting in the cloud would cost us £11,000 a year.
At the moment if a user wants to print a large document, it goes to the internal print server and then released by the user at the printer - takes seconds.
Moving to the cloud would send that data out of school, to the cloud and then back again.
One school trust, which has just migrated its primaries completely to the cloud, says this takes around ten MINUTES for a large print job, and they're having to put the print server back in-house.
They want straightforward reproducible systems that work - like ERP/Finance/HR that can scale from parish to County/Region Council and interface to mainly the same external stakeholders.
All councils - even devolved - basically do the same things.
Pandering to local choices/processes is baking inefficiency, complexity and is the antithesis of change delivering value.
See Birmingham City Council, West Sussex Council and related organisations repeated IT failures and Oracle or SAP.
Aren't companies starting to move away from the cloud as it makes them less resilient, less secure, and places them at the mercy of price hikes?
I'm guessing this government policy was cooked up some time ago, governments usually being last on the bandwagon. Perhaps in tandem with this, they need to begin work on a future strategy to move these folks back off the cloud, in a few years time.
Or maybe they could just keep them off the cloud, improve their security, and donate £1bn to Oracle, compensating them for the public money they would otherwise have bagged.
They'll buy into the myth that the Cloud has all the security specialist the Government is not willing to properly pay for unless they're called consultants, and then they're about 3x more costly than a properly paid consultant would have cost them. But then there wouldn't be backhanders, I suppose.
I love cloud, it's great but government's rush to it worries me.
First, it isn't the lowest cost unless you design & use it carefully. Second, it's an eggs in one basket situation regardless that they have more than 1 datacenter in the UK. Third, the companies are not British and we do not know what happens in extremis. For example, it is entirely possible that we enter a hot war with Russia that isn't nuclear. What happens if all undersea cables are lost, satellites being destroyed, what happens if we break with the US, they certainly wont be helping if they have their own problems. Sound absurd? Weird stuff happens in times of chaos. I commented on here that we need to spend more time developing manual disaster recover and it got a lot of down votes. There are already many cyber attacks, what if they were just testing and probes and multiplied by 10 or more with many systems were disabled? As far as we know the banks have proved resilient, that may change, imagine your credit vanished and cash machines were emptied in a day not to be refilled?
We need to be more diligent in planning for the unthinkable.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
