Slashdot Mirror
Cloud-Sourcing's Long-Term Impact On IT Careers
snydeq writes "InfoWorld provides a reality check on the impact cloud computing will have on IT jobs, the overall effects of which will likely resemble those of outsourcing, automation, and utility computing — in other words, a movement away from the nuts and bolts of technology toward the business end of the organization. This shift from 'blue-collar IT to white-collar IT' will be accompanied by greater demand for IT pros experienced with virtualization and Web scale-out deployments, even among midlevel organizations, and greater emphasis on SaaS integration among in-house development teams, analysts say. And though the large-scale impact of 'cloud-sourcing' is likely a decade away, those not versed in vendor contract management, cloud integration, analytics, and RIA and mobile development may find themselves pushed toward the less technical jobs to come, those that will require days full of conference calls and putting out fires caused by doing business in the cloud."
While "Cloud" computing is here to stay, it is a while off, and opens interesting problems with different legal requirements caused by regulations. How does SOX integrate with Cloud computing. As it stands currently, it really doesn't because there is a lack, or at least perceived lack of accountability. While I can agree that we will see more widespread rollouts in the next ten years, for those working at publicly traded companies or in Health Care, you can expect to see a lot of systems still in house, albeit virtualized and distributed, rather than off site. All this article really does, is reiterate the importance for any IT professional to stay current or move into management.
The musings of just another geek and his junk.
And I say that as a sysadmin myself.
There are thousands of businesses around the world that are just large enough to need a couple of full-time IT people of some description. These businesses account for a lot of IT jobs. If you actually crunch the numbers, the huge companies of this world don't employ that many people as a percentage of the working population.
SaaS allows a business to grow much larger before it needs a full-time IT presence than was previously possible - all those crappy little applications (of which there are thousands) that IT technicians, sysadmins etc. got to know backwards and inside out and were next to useless in their next job are going the way of the dodo. PC so full of viruses and spyware it's virtually unusable? Considering the amount of money you'll pay per month for a full-time member of staff, it may well be cheaper to keep a couple of spares in the cupboard and just bin it when you hit trouble.
This is great for the business - they can get more done for less money. Not so much for the sysadmin.
1) Flexibility: difficult to mold SaaS solution to your specific business operations.
2) Reliability: requiring a connection to the internet adds an additional point of failure.
3) Speed: easy to get 40mbps internally. Internet connect is more likely to be 1.5mbps split 50 ways.
4) Cost: from what I have seen, SaaS is not especially cheap.
5) Security: debatable.
6) Vendor-lockin: if you need something changed on the server side, you only have one choice for the developers.
I don't really know, and I suppose a lot of it is situational, but I am not certain that that is going to take over the world any time soon.
What the Industry sometimes fails to realize is that it is IT people who make or break the products. From the majority of /, readers responses to all these Cloud Computing posts, the main concerns are reliability and security. Reliability may be solved soon, but I feel security will always be a neverending list of crackers and incompetence on the part of the cloud utility. Too many stories of losing usb keys, laptops, security passes, passwords, etc, on the part of large "no fail" companies that should know better. Most businesses will be very very adverse to giving up control of their data, and somehow I don't see that ever changing, even when they claim the risk is almost 0%.
When the cloud utility is Google, IBM, Amazon or some other mega corporation there is no way in hell that the security or reliability will be any worse than what a small company can muster on their own. The writing's been on the wall for a long time now, but many techies that love dealing with server infrastructure doesn't want to give up control of what they have, or aren't able to admit that someone else can do their job better. It is the same thinking which is why some people still insist on building their own servers by assembling the hardware components themselves and running Gentoo instead of buying something pre-packaged. Economies of scale doesn't work out very well for those.
Football Odds
I would beg to differ. Most don't see the NEED for virtualization, but I can assure you that it is very much a hidden asset (need) that is not being fulfilled.
Most people see technology as "we'll fix it when it breaks". And they have no idea how much their lives and businesses depend on it, until it disappears.
Long term replacement plans are never employed, and businesses end up being held hostage by disaster.
I know of plenty of small businesses who have gone out of business because their ONE server has died, and they can't recover the data. Their one server died, because it was ten to twelve years old, and they never thought to do backups, or replace it, or whatever because they don't have a full time IT staff dedicated to supporting it to explain why they need to worry about such things.
Virtualization is just another means of removing Hardware dependencies from the equation. That Abstraction layer is such that you can migrate the server to new hardware faster and easier than trying to migrate it. I can do a whole new hardware platform migration in a few minutes with VMWare. Shut the old machine down, move the VM to the new box, fire it up.
And when you add in features like snapshots it just rocks. I can't tell you how much snapshots are worth. I've even used VM snapshot in a criminal investigation, where we could show exactly when the fraud was started.
It saves them money, and headaches, and eventually it will save their business.
Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
There is more to SAAS and Cloud computing than the technical aspect. People, in general, are averse to change, and introducing another technology into the Peter Principle must be factored in as well. As our cultures become more complex and risk-averse, that is more enhanced, meaning that there has to be a very tangible REASON to change that is worth the upsetting of the status quo.
So for the short term, only early adopters are going to dip their toes in this water, by putting hot-swap sites on a cloud (though the reliability of the cloud service still needs some upgrading). The mid-term will see server hosting on the cloud, using the same applications and processes they would have anyway. Only when SAAS can be presented as a viable economic model to the companies paying for it will they look at it.
I know some of the guys at Amazon who are working on this. Without giving any secrets away, there are some technical and security issues that will take some re-architecture to overcome before client/server software could be reliably and easily hosted there. Amazon moves as quickly as anyplace in resolving such things, but it will take some time.
I think you're right that IT people make or break a solution and the same is true of cloud models. The need for IT people to design, build and repair the automation will stay in the new model. I'm not sure if this is "blue collar" IT or not (sort of a dumb concept). I wish cloud wasn't pushed as a new concept because in reality it's a convergence of a number of concepts that have been in the industry for some time. I think that's what makes it a little more of a reality than the traditional new buzzwords. Cloud Computing is taking the standardization efforts of SOA and ITIL and adding automation and virtualization. Since most of us now have IAAS and PAAS documented it makes it easier for Cloud to catch on.
There will be a shift in roles. People that today have most of their job defined by Server Build, Application installation, or OS installation are in danger. They need to either script themselves out of a job, thereby becoming "Automation Experts" or someone will likely do it for them. Current timelines within organizations for server build of 1-2 weeks is going to become a pressing issue as we exit this recession and try to ramp up new projects. Cloud is available at the right time to address those concerns and is based in concepts that people have already been using for a couple of years. It's possible that the word Cloud will go away but the concepts that it brings are going to stick around for a long time, and be a game changer.
This signature would be better if I was creative.
What cloud computing will do is to erase the advantage that one company has over their competitors. Take a company who has a 'better' business process and has built their in-house IT systems around it. Now, have them move to a cloud solution. The first things that must go are their custom data models and algorithms. Companies re-engineer themselves to fit OTS applications much more often than they customize the apps to fit the business. The next thing to go wil be their in-house development staff. No sense in keeping them around when nothing is left to be written that is specific to their company. Next, the executives who have domain knowledge specific to that companies' core processes can go. Since everyone in the industry (or across many) use the same processes, why pay big salaries when the supply of suitable talent is much larger.
So tell me: Why should I do business with your company? What do you do that your competitors don't? I mean, the cheap ones, in India or China.
Have gnu, will travel.
Basically a high paying job doing fun, researchy stuff is an oxymoron.
... Go back as a grad student (with 0 responsibility IMO), and forget about making $100K a year.
Want to work with or develop the latest tech, coolest languages, ingenious scalable configurations?
Otherwise, as the OP mentioned, it's business as usual, suck it up.
Not to mention this is another article in a long series of articles that essentially says "do your bosses job for less money than they will ever make or you will be fired!" type bullshit that managers just eat up(without even being remotely critical of the logic or conclusions). Essentially it allows managers to use fear to get their underlings to do even more work so the manager can spend more time golfing while getting a higher salary with the belief that only management is indispensable, but honestly I think its the opposite. If the peons really do the work of management in addition to their own they are eventually going to figure out how to start leaner companies without the need to feed a small number of incredibly useless, but well connected, people massive amounts of cash and management will be out of a job...
:P
Well, at least my hypothesis is at least as plausible as the one presented in the article
Monstar L