Banks talk big cloud game but few have migrated over 30% of apps

A report into cloud adoption in the international banking industry shows that despite a broad appetite for cloud services, only around a third of banks have migrated more than 30 percent of their applications. The Future of Cloud in Banking report found the majority of retail and commercial banks aim to triple their use of …

[Scott Broukell]

Well, I suppose it is the way of the furure. Personally I would be happier if I were fully confident that my bank had thoroughly tested and resilient pen and paper (and pocket calculator) Backup Plan for when the cloudiness goes off-line for a bit of a rest, as it were.

I mean IRL clouds do have a tendancy to burst etc. Fingers crossed everyone.

Just saying.

[ManMountain1]

Santander recently claimed 80% of workloads were in the cloud but they were classifying on-premises private cloud in that number. I'm sure there is a place for multi-tenant / public cloud for banks but hopefully they are smart enough to keep the really important stuff under their own control!

[NeilPost]

I’d personally prefer for my Bank Info not to be with a Cloud Provider, even single tenant/private compute services.

Same for all manner of other GDPR areas. Health, Financial, Legal, Personal, Tax etc.

[Anonymous Coward]

Indeed.

When I read that about Santander, that was a big red flag for me too.

[Gene Cash]

Which ones aren't using cloud?

Let me know and I'll move my accounts over to them!

[amanfromMars 1]

Damned if they do and spurned and overcome with bank runs if they don’t

Come into my parlour say spiders to flies ....... as does future business to current parasites .... is the present dire straits state of monetary state play and federal reserve bankrupting shenanigans.

[Kevin McMurtrie]

Eventually (manually) consistent?

Cloud hosting is generally a bad idea for "exactly once" operations. Frequent minor outages have you tracing down obscure fault recovery bugs forever, and each incident must be resolved quickly and perfectly when you're a bank. Self-hosted can easily go a year or more without a single unplanned outage.

[FozzyBear]

Or just maybe you have to ask

you look at the legacy insurance, superannuation and banking product applications that have been happily whirling away for years/decades running on old iron without an issue. You then look at the cost of porting/hosting it as a cloud service and a very reasonable question arises.

What's the benefit again?

[Warm Braw]

Re: Or just maybe you have to ask

If only that were the case and we didn't have myriad examples of failed or duplicated payments, systems being unavailable for days and general precariousness. Most legacy bank systems are cobbled together from multiple layers of technologies of different vintages and, often, with only distant memories of design criteria.

That's not to say that I'm advocating for the cloud - the old systems are fragile enough as it is and probably wouldn't survive the transition given all their undocumented quirks.

The "fintech disrupters" are the ones that have, quite reasonably, embraced the cloud as it means they don't need the legacy infrastructure and can scale relatively painlessly. However, as soon as they have their own technology legacy to deal with they're going to be trapped in exactly the same way. I'm not sure many of them have sufficient of a business model to ever get to that point, mind, but that's another story.

[Shepard]

Well of course

Some banks still have ATM fleets running Windows XP on Pentium IV Celerons, most of the rest of them are stuck on 32-bit Windows 7 (usually not even patched up to the last January 14, 2020 patch, let alone having extended license to continue receiving security updates after EOL date). Small number is only now migrating to Windows 10 1607.

The reason? Cost.

They don't want to soend money on new PC cores capable of running modern software, not to mention all the recertifications they'd have to go through with their processing houses.

Is anyone really surprised they are not keen on buying into cloud stuff?

[BOFH in Training]

What do they mean by "apply advanced analytics" in the cloud?

Can't you do that no matter how you host your data or web apps?

It's just a matter of logging whatever info you need and doing whatever analytics you want, from my understanding. Be it just web access logs, or some other more advanced forms of logging. After all you can log pretty much anything and everything you want, if you are self hosted. I will imagine there will be less access to logs in a cloud provider - you being pretty much dependant on whatever they give you access to. And dependant on whatever analytics they provide you.

[pavel.petrman]

This is good news.

See title. I do hope the 30% of workload comprises of bonus calculation and allocation, senior management vacation time and pension funds and perhaps facility leasing. If I find that anything to do with my money has been offloaded to the cloud, I'd have to have a stern word with the regulator!

[Anonymous Coward]

When you have 1000s of apps

Banks tend to have 1000s of very small custom apps doing stuff that isn't going to change: "print a contract for a customer", etc.

By the time a developer has figured out what the code does, changed it, moved it, pointed GUIs at the new version, got it tested, told people who have to use it once a year what has changed... All with the job satisfaction of having basically changed nothing that anyone really cares about... you realise why COBOL compilers still exist.

That's before you appreciate how much cheaper it is to have a few rooms with rack mounted PCs running KVM... Outsourcing to clouds is expensive...

[spireite]

Re: When you have 1000s of apps

Leveraging the S in SOLID, before SOLID became something widely known in many ways!!

I'm still on the side of the fence that says for many things Cloud is cost too far.

Has its places, but some industries shouldn't be going there. A bank is one of them!

[pavel.petrman]

Re: When you have 1000s of apps

The last time I worked for a bank almost two decades ago, even then they utilized their infrastructure the way today called "as a service" - huge machine with many processors and big memory, both enabled and disabled remotelyby vendor on on-demand basis. The requirements, even then, fluctuated wildly. One week the whole beast was fully booked, the next hardly four processors and a gigabyte memory was booked. Much of the work was doneby external consultants (the likes of Accenture).

I, for one, wouldn't want to be responsible for coordinating all this with the bank's own server room both technically and financially. Yes, the processing power, memory and storage is very expensive in cloud, but I wouldn't bet on it being more expensive than how banks operate outside clouds.

[Anonymous Coward]

Still forced to use IE

My organisation uses Pay.UK - their blurb says

"In 2019 Pay.UK became the home for the UK’s retail

payment schemes Bacs, Faster Payments, Cheques.

By bringing the schemes together as Pay.UK, we

put the needs of consumers and businesses at the

heart of everything we do. As one operator for all

UK retail payments, we make sure that every retail

payment sent or transferred in the UK is done

safely and securely."

Yet their online banking transaction software still uses IE and we're supposed to keep applying workarounds and have to do so every 30 days.

[Dan 55]

"Less than half said business leaders in their bank understood 'opportunities of cloud'"

Perhaps they do but maybe they also understand the 'disadvantages of cloud'.

I'm feeling optimistic today...

[Anonymous Coward]

Skewed report

The problem with this article is that the report it’s based upon is sponsored by a cloud services provider. The report is hardly going to say that cloud isn’t the right choice?

[Anonymous Coward]

A question from the ignorant

It strikes me that there are two possible ways forward - Have I missed anything?

Firstly. . .

The challenger banks are busy making fresh new systems to run on modern technologies (nothing wrong with old systems other than at some point all things die). In a few years when these systems have proved themselves the challengers will be ripe for takover, and the heritage banks get a running, proved and certified system to migrate customers onto.

Alternatively. . .

A few banks could form a jointly owned Cloud provider focusing on building a modern system with security at its core. A 3 way data center (say London, Amsterdam, Madrid) should be resilient if one burnt down, bombed etc.

Either way they need to own their own destiny.

Am I wrong?

[Anonymous Coward]

Re: A question from the ignorant

Just thought, so I will answer my own question. . .

BREXIT

[dinsdale54]

Report produced by cloud provider disappointed that not everybody uses the cloud for everything. News at Ten!

I mentioned this previously but a bank I used to do a lot of work with said that for fixed demand, the cloud was just hosting which they could do much cheaper than any of the cloud providers. They already have many large datacentres all over the world and were running at way lower costs than any of the cloud providers.

Another customer in Australia moved a huge amount of their SQL Server workload to the cloud - successfully - and then discovered that the bills made them the biggest customer in the country. They moved most of it back on prem to save money.

Cloud services are very useful to many people and organisations. Not all of them..

[A. Coatsworth]

"Less than half said business leaders in their bank understood 'opportunities of cloud'"

... And less than half these opportunities were thought out less than half as well as they deserved.