There really are technology stacks and business use cases that should never be moved off of big iron. Then there’s the rest of us. Chances are very high if you’re not already starting to move your data storage to the cloud, it’s being discussed right now in your business. I helped put together a new report on the current state of databases that are not under your direct control. It’s free to download and available over at Dzone.
If you haven’t made the move yet, or are just getting started on becoming one of the millions of companies that will be hybrid, part cloud-based, part local, then I strongly recommend you check it out.
The trends are clear. Smart people are taking advantage of what services like Azure, AWS and Google have to offer. This is especially true in the Platform as a Service (PaaS) space. The technology is getting better and better. You can do things with it that only gigantic corporations ever dream about. Please, go and read the report then come back around and let me know what you think.
Yes. The cloud should serve as an optional tool among many others and not the sole end point for everything (of which data is included) and yet that’s what most are pushing for with little to no regard for security/redundancy then for speed/cost. There is not enough healthy skepticism of the cloud. Its being treated as a given technology with few if any risks. Like the internet of things and streaming technologies, the Cloud is being marketed as and treated as a fail-proof technology that is the digital savior of mankind.
To use a classic phrase, its foolish to put all your eggs in one basket and yet that’s what is happening with the cloud and the other technology cloud success depends on which is just as equally bad to rely solely on, streaming. Its not that the cloud is bad but that placing everything into it and relying solely on it along that is very short sighted; a problem just waiting to happen. Instead of asking if the cloud is the right solution to some problem the approach has been to figure out how to market the cloud as the solution to every (technology) problem. When every problem looks like a nail then a hammer is the only tool you need.
The race to the cloud has succeeded because those behind it were smart enough to push the cost angle and get buy in from business. Once you have enough customers moved to your cloud based solution its easier to nudge some more and then finally force the remaining holdouts. The cloud is a race to the bottom, the question is how many years before the cloud crashes so badly that enough realize how foolish it was to place all their bets on it.
I try hard not to argue with people online, especially on my blog. However, where on earth did I suggest above, or in the article, that all the eggs should be placed in a single basket? I explicitly state, quite clearly, and I quote, “…just getting started on becoming one of the millions of companies that will be hybrid, part cloud-based, part local…”. I’m not the greatest writer and I’m not terribly bright, but I’m willing to put solid money down, that sentence does not mean “move everything to the cloud.”
You want to be critical, fine. Please, at least be accurate.
It looks to me like there is a miscommunication + a misunderstanding here….
I took the comments of TheCloudIsNotTheDigitalSaivorOfMankind to be generic and not actually directed at you (GF).
Many of us are being bombarded by endless projects led by morons who read the same crap from the same place (CIO Magazine(?)) that the cloud is the be-all and end-all of hardware cost savings when it has been proven more than enough times to be the exact opposite. I didn’t see anything there that criticised your post – it was just at a tangent to it 🙂
Funny how your article mentioned ‘Hybrid’ as a solution which I think is a valid way to go yet, when dealing with the cloud vendors, they are the ones who say that “hybrids are bad…” (in a Mr Mackey voice (South Park)).
I think cloud technology has a place and adds value but it should never be used for those databases that will see you on Page 3 (or even Page 1) of a major newspaper for a data corruption or data being stolen and is picked up by politicians. I say to use it for consolidation of instances (reduce sprawl) and to house data that isn’t going to present a security or privacy risk to anyone.
You should go and read my article in the link then. I talk specifically about data in the cloud and whether it’s more or less safe than locally hosted data. Specifically talking about the most cloud-focused data stores like Azure SQL Database, CosmosDB, RDS, indications are, overall, they’re more secure than locally hosted data on average. I go into detail why in the article. This is the kind of thing I’m trying to fight against. “You have to move everything” “It’s less secure” These are myths and misconceptions.
We are 100% cloud and let me tell you, it’s a brave new world. If a vendor tels you to plan for failures, by all means please plan for failures. They will happen.
Test, test, and test your infrastructure design decisions in the cloud. We don’t useany of the managed services, so we went through about four configurations before we found the best i stances for our SQL Server workload.
I could go on, but it was a long night of on-call duties.
Yikes!
Very much matches reality though. Nothing truly magical about the cloud. If anything, it reemphasizes the need to get the fundamentals right.
Outside of people running a website, servers for an online game or some other thing where multiple redundant servers in multiple countries is a huge plus point, I really don’t understand why anyone would pay *more* to have their data stored off-site.
I can buy a server for a few hundred bucks, install a copy of Windows Server and SQL Server on it and I’m done. sure, I have to pay an annual licence for Windows and SQL Server, but that’s orders of magnitude cheaper than hosting it in Azure or AWS.
If my internet connection goes down, I still have access to my data. If Microsoft or Amazon get DDOS’d, I still have access to my data. If I forget to pay my subscription, or my company starts having cashflow problems, I still have access to my data.
Most cloud implementations I’ve seen have an entire physical installation as a failover, “just in case”, which basically doubles the cost and effort to keep it running. When you could just run on the failover anyway, and bin off the cloud…
It’s a lot of marketing hype for very little benefit in my opinion. Sure, the processing time might be quicker, but unless your WAN is as fast as your LAN, which for the vast majority of businesses, it really isn’t, then it’s always going to be slower to use.
There is more to it than that. It’s really not about picking up what you have locally and moving that to the cloud. Some people do that, sure. However, that’s not the interesting stuff, and certainly not the things that are changing paradigms. Doing the same thing somewhere else that you do locally isn’t a change at all. No, I’d suggest digging further into the Platform as a Service offerings. That’s where interesting stuff is occuring, and it’s more than what most businesses can do locally.