I believe the earlier name for "cloud services" was timesharing. The 70's called and want their VM370/TSO back.

You might look into retrospect (http://retrospect.com/). The have clients for macs and PC (and some flavors of Linux) and it's pretty easy to use. You can back up remotely (on schedule or on demand) and could restore locally of the hard drive. You & your neighbor can also back up locally onto a 2nd hard drive. The program has been around for 20+ years, it's reasonably price and the support is slightly above average. They have a free trial.

I have to agree that the host/server/bandwidth costs should be a relatively small factor on your calculation. Reliability, security and responsiveness really should be more important. The difference between top tier and bottom tier hosting/cloud is probably no more than a factor of 2 -- you can easily burn thru that savings with a couple of hours of downtime or a hosting vendor screw up.

If cost is really important, I'd get it working first at a top tier vendor and then overtime try to squeeze out costs--either negotiating a better rate (based on your volume) or switching to a lower cost vendor.

Alternative, why not just buy more bandwidth to your location. The bandwidth costs should be relatively low compared to the overall project costs. Also, this will provide you with office redundancy (at least at some level).

Too often in trying to save money, people focus on the wrong part of the problem.

I have to say with all of the big names having problems recently this has been one of the best weeks ever for the lowly corporate sys admin. Now if the company's email, file or web server--or even the coffee machine goes down, they can point to the big names that also have problems. It's great to be able to say that even at companies like Amazon, Google or Microsoft with all of their talents their servers also have problems. It's the greatest excuse ever for tripping over the power cord. And if that doesn't work, you can always blame the NSA for the typo in your email or the late TPS reports.

Thanks everyone and happy SysAdmin day! (which isn't today, but due to the unexpected outage is running late)

How up time is calculated is one of the really weaselly ways that companies set up SLAs. Some companies don't start counting downtime until it's reported, others require a minimum threshold of downtime before it counts, others define available in somewhat meaningless terms (e.g., server up, but network down doesn't count).

Somewhat OT, but I was pleasantly surprised to find a number of third-party sources to repair micro USB connectors. I don't know if it was a manufacturing issue, but the micro USB went on my Samsung & HTC at about the same time. For around $30 to $40 each, I was able to get them repaired.

As long as what you want isn't iSCSI, FCoE or 10gig either.

I'm waiting for someone to announce a rack mount kit for a tubular computer case. Actually, anyone who needs cheap computing horsepower probably isn't buying mac computers. These are targeted at the single-shingle video producers and others for whom a single "work station" is fine. At the high end, all of the disk is iSCSI or something similar so TB doesn't matter.

At least on the Prius once the car is running even if you move the key fob out of range, the car keeps running (actually a good safety feature as you wouldn't want the car to shutdown on a key fob failure.) On the Prius (and maybe other Toyotas), there is a metal key for mechanically unlocking the driver's side door and a electronic slot for starting the car. You can use the electronic slot if the key fob batter is completely dead so I suspect it's a passive NFC device. There is also a mode that you can disable the active detection feature and always have to use the dashboard slot. Other models probably have similar features.

Other than the looks, you don't really need the MBP sitting on your desk. You could use any commodity laptop to SSH into the server farm. Nothing wrong with the MBP, but overkill as a dumb terminal.

In the late 70's I worked on one of the original Crays and it had an option to specify a file deletion date (or retention time) when you created a file. The file would be automatically deleted (or maybe archived) at the appointed time. I've often thought that this would be useful in a desktop OS--when I create the file, specify that it should be deleted in 2 weeks. Same with email--it would be great if you could read an email and then indicate that its retention should be two weeks or one year... and then it would automatically disappear.

The approach proposed is that it teaches you a trade. The problem is that you will likely quickly cap out on salary and opportunity. Much like a plumber can get a license in a year or so and start making 30-50K... that's pretty much the max (unless you start your own business).

In the tech field, there is always another kid coming along with more current skills and willing to work at a starter salary.

Ideally, a college education teaches you how to learn... not merely a trade.

If you're looking to learn a trade, the 10-week "truck driver training school" approach might work.

Of course, there are the few rare exceptions where a non-college graduate has gone on to great things. But for the vast major of people a good solid education is more likely to equip them for a lifetime career than gambling on starting a hit business.

Although the original posted pointed out a handful of successful non-college graduates, I'm guessing that there are millions of non-college graduate failures that you've never heard of.

Actually the buggy whip business isn't dead, but has turned into a niche market. A quick google search revealed http://www.jedediahsbuggywhip.org/sales.nxg which goes after the accurate period reproduction whips and repairs and has been in business since 1851. A different company has gone after the modern market with LED buggy whips (for visibility at night). The advantage is that these stores can reach a national market from a centralized location (much like Netflix).

The real solution is to redefine the business using the existing customers as a base...video game rentals, snack food/beer with a side of video. But it's a pretty tough challenge in a saturated retail market with not a lot of IP other than a customer list, knowledge of movies and location.

I believe hardware is done by an electrical engineer.

Unlike many other databases, errors can be tolerated in facebook. If a post gets lost or a connection or two dropped it really doesn't cost Facebook anything--and it's unlikely to be noticed. And downtime and retries are tolerated by the users.

  Try running a real-time, financial system like credit card authorization & processing (which probably has more than 1 billion users), needs to balance at the end of the day and has response requirements measuring under 250 ms.

  Facebook is just better at promotions. There are other databases that are bigger, have tighter response requirements and are more complex. It's all about buzz.