After figuring out that hope is not a strategy, SAP has a new one: We're gonna shift on-prem customers to the cloud! Hope is not a strategy, and that common aphorism is coming back to haunt enterprise software outfit SAP. In the first quarter of the year, the business was keen to get investors to understand that, despite the adverse effects of COVID-19 on the global economy in the short term, stalled projects would come back on stream. On a …
If you'll excuse my French, but what is it that they say ...... You can't polish a turd but you can cover it in glitter ...... and such is always a hard sell when losing a cash cow.
It is surely cold comfort, but SAP is not alone struggling to make great book in a particularly and peculiarly competitive field that values prime timely access to your intellectual property.
Perhaps - like Oracle - for ERP implementations - they should work harder with their partners and customers to make it easier to implement without major trauma and lead to fucked up, under-delivered and failed implementations.
I think the bot is getting better, but there's some anomalous grammar issues. Besides, Myth Busters proved that you can polish a turd. They did it on the show.
https://go.discovery.com/tv-shows/mythbusters/videos/the-buster-awards-polished-poop
One of the many reasons for NOT moving to the cloud is security. Using other people's computers (ie the cloud) makes it far more likely that information will be inadvertently disclosed. Add to that most long term cloud computing costs far more than using ones own hardware. (Look at the AWS charges for the equivalent of a small server and compare that to the cost of the same server - then work out how many months before AWS is the more expensive option.)
Cloud reliability is going to be another, as the cost of downtime in an enterprise ERP system is pretty horrific for some people, and getting a canned "The cloud is down" response is cold comfort. At least with a self administered deployments, and I use that term as many of these outfits are managing their own cloud deployments, so "on-premise" seems misleading, you can buy and ensure the number of 9's you need, and staff it with people who are aligned with YOUR needs, not SAPs.
Anyone who thinks their on prem systems are more secure than cloud systems is fooling themselves. If you are connected to the internet in any way, you are vulnerable & a target.
The difference between cloud & on prem is that one has hundreds (thousands?) of engineers solely focused on securing systems and the other has maybe a dozen. I'll leave it up to the reader to figure out which is likely to be more secure and who has the experience/resources needed to fend off the most sophisticated rogue & nation-state attackers...
As far as cost - it's all about application architecture, the cost is not 1-1 as, in a properely architected cloud application you only pay for surge capacity when you need it instead of 24/7/365....
When a Dog see's Another Dog, It Knows the Other Dogs Protocol.
As the Other Dog's Protocol is its Own Protocol.
It begins to Bark and Make itself Look Bigger as it Knows What Big P1g does to Little P1g.
As that is What it does itself.
Corporations will Not put their data on the Cloud, as they Know What will Happen.
You've told yourself a Lie and began to believe you own Bullsh1t.
You Yourself are the Virus. It's just that you've began to believe your own bullsh1t that you're an Angel.
Two problems. Physical access. Logical access.
You will likely have a difficult time getting in my data center. Aws has a hundred or so, Lot's of people I did not vet touch their gear, so yea, they NEED more security.
Logically, if I create a path from the internet to/from my AWS environment how is that any different from a security perspective?
Lastly 'properly architected' applications. Bwahahahahah. Umm. Designing elastic applications for cloud, is decidedly non-trivial. Maintenance of same is decidedly non-trivial. You are speaking from a position of ignorance. I got 100 or so machines in AWS, cause the c-suite wants it. None of it scales much, but at least it's expensive. Talk to me when you have a fully resilient app in aws, lambda triggers and control/monitoring set up. It's slick but designing a system that magically adds nodes to a service cluster is intense. when the function can be accomplished with 2 guest's in 2 data centers, for redundancy, why am I spending months architecting?
Um...sorry, wrong answer. The security personnel are only there to make sure the hardware is secure from unauthorized physical access and to make sure that the base operating system of the server is secure. They are not going to work to keep YOUR application secure. That's on YOU, your software development team, and your vendor.
<<Look at the AWS charges for the equivalent of a small server and compare that to the cost of the same server>>
I've never seen a cost comparison of cloud vs. on premise that, in addition to the usual server HW costs, factors in the costs of: real estate space (data center), proper and redundant cooling, redundant power sources, redundant network links, fire prevention, access management, support personnel, backup rotation plus storage and medium costs. And on top of that, the monitoring of all that, 24/7, please.
If you compare just the server hardware costs, sure, on-premise is cheaper. After including all of the above mentioned costs, you realize that economies of scale favor having huge data centers, and in that regard you (or I) simply cannot compete with the likes of AWS, Google, Azure, etc.
If your costs are smaller including all the additional items, you're just being foolish by not setting up your own cloud business and competing with the big providers. If your costs are not smaller, please shut up. Which is going to be the case in 99,99% of the times.
And you hit the nail on the head. Hosting your own environment is way cheaper than hosting it on the cloud. Hardware is a one time investment (barring failures and part replacements), you still need to pay for internet connectivity, you still pay for software licensing fees for your software. The added expense is the operating system (usually a one time cost unless you opt for a support contract), electricity to run the server, backups, and a few bodies to make sure the server is running well and run patches into it.
Most medium to large businesses have their own IT staff anyways. Looking at the TOC vs ROI between the cloud and hosting it yourself, the long term is that hosting it yourself is cheaper. Not only that, it's generally considered to be more secure. Just look at how many 'open' databases have been found on the internet with confidential data for the taking.
Granted, when you go cloud, you are paying for the administration of the physical server itself, not the virtual image that's running on it. But you would pay that cost anyways hosting it yourself.
<<Hosting your own environment is way cheaper than hosting it on the cloud>>
Ironically, AWS started when the Amazon folds asked themselves "what if we rent this equipment that we're seldom using?"
So if it is true that hosting your own environment is cheaper than the cloud at the same levels of quality of service and reliability, why on earth aren't you renting your spare capacity? You're either being a fool for losing money, intentionally not comparing apples to apples (that is, ignoring all the true costs involved) or just... plain lying to yourself.
Same thoughts here. Same problems, same challenges with, let's call it, sensitive and very definitely PI data. So: no way our company (and in fact the rest of the related businesses we are working with) are going to do that. Heck, we even have our own vidoe conference servers (two different plattforms) and are still on the non-coudy MS office stuff. Considering the size, scope and architecture of the whole kaboodle I would actually classify us as a small cloud provider. Which is probably not what SAP and Atlassian have in mind.
I think SAP stock tanked as hard as it did because of internal problems and what SAP is as a company. They have been accumulating technical debt for years and papering over the cracks in their quarterlies with predatory sales practices. This is just when the house of cards started to fall. SAP has ruined its reputation in the industry by over promising what the platform could deliver, intentionally causing cost overruns as a sales tactic, and have failed to resolve its performance issues for decades.
The fact that SAP's stock cratered when triggered by broader market forces does not mean other companies are at grave risk for collapse. SAP was just teetering on the edge of the abyss and got nudged over. The only risk SAP represents to other companies stock price is when the make the mistake of switching to them. There is probably a viable hedge fund strategy there, shorting companies that are doing SAP rollouts.
... and you wonder if it's just the usual techy snark.
But only 8100 "completed" deployments out of 15,100? S4/HANA has been around for 5 years, and only just over half of the deployments are live?
Those poor, poor companies that were sold such a massive pup, and are still not able to use the hugely expensive software they paid for.
ERP disasters are caused by the process of adopting ERP.
ERP is not just just software, ERP is about business process.
If your business cannot adopt ERP methodology, guess what, your ERP software isn't going to work.
In fact, in adopting ERP, the business change is many times more significant than the IT one. You are absolutely heading for disaster if you treat an ERP adoption project as an IT one.
That's where this has gone wrong for people ever since someone tried to emulate Toyota with MRP.
From my past experience with a MRP2 software vendor, I'd say about 10% of customers get it right. You need to fully embrace it and listen to the business consultants (that you have to pay for). If you cannot follow their advice then it will fail EVERY TIME.
The problem usually is that people think its just a piece of software and magically its going to make their business better.
""This time, we'll be moving large parts of our ERP customer base from on-premise to the cloud, ..."
Wow, so a SW vendor is just gonna be "moving" their customers to cloud, like you move a pawn on a checkboard ??!!!
Like said customer/pawn is just a simple piece of wood/plastic too dumb to even object ?
SAP people need to put their arrogance back in the pandora box, really.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
