It definitely looks as though they ended up having to manage it themselves.
They're not "manag[ing] it themselves" -- it's a gigantic stretch to say that software RAID on an instantly-configured EBS volume is remotely equivalent to specifying, ordering, building, installing and maintaining a number of RAID arrays in a data center.
Basically this seems to put a fairly low upper limit to using 'the cloud' for something a little larger before you get back exactly the same kinds of issues that you were dealing with when using self-hosted hardware, only at a higher price point.
This is ridiculous. The primary issues with self-hosted hardware:
- Energy usage. Data centers only have so much power and cooling to spare, what's available costs money, and you constantly have to monitor/juggle utilization and search for more energy-efficient hardware, which leads to the next issue ...
- Physical plant. You often need to buy it well ahead of time to ensure that you'll have power and space available for expansion should you need it. You need to wire that space, buy switches and routers, and that leads to the next issue ...
- People. You'll need people to order, build, and install those servers. You'll need people to swap out dead components in those servers. You'll need trained people to install PDUs and console management systems, to configure and run your switches and routers.
EC2 and similar pushes these issues upstream. You can manage your servers in software. You don't have any routers or switches to manage, you don't have to order, build, or install hardware. You don't have to expend significant capital outlay on servers or rack space.
I probably am not getting my point across very well.
Pushing issues upstream does not make them go away, it makes them someone elses problem. If that someone else doesn't take care of those issues then you end up having to solve them yourself.
And I've played enough with software raid that I know that configuring it to perform well is not a walk in the park, in fact I think it may be harder than getting a good hardware raid solution up and running on a dedicated box.
The nice thing of course of an EC2 setup is that once you've figured out how to do it you can do it again without much trouble.
As for the energy usage and the physical plant, that is entirely up to your way of using your servers. For instance I try to balance the quality of the service with the load on the servers to gracefully degrade the service when the load is at its peak (which is only a few hours per day anyway). That way I maximize my flat-rate payments on bandwidth and serverlease at a relatively small fraction of what it would cost me to get similar performance out of the various cloud suppliers offering. In a cloud environment those servers would be using just as much power and AC as they do today.
It takes me a little longer to get a server provisioned, on the order of 2 to 3 days, but that is 2 to 3 days, whether I order one, 10 or 100 servers. Very few businesses would ever need to grow faster than that.
Maybe my business is a lucky one in that it can make optimal use of a dedicated server setup but I see plenty of people choosing for a cloud based solution when if you run the numbers it makes very little sense.
The cloud comes in to its own if you have wildly fluctuating loads and/or jobs that need large numbers of machines for a relatively short period.
But for the majority of longer term high bandwidth uses I can't make the numbers work at all.
- What vendor you've found that can provide you with 100 leased servers in 2-3 days while costing less than a cloud provider and not requiring you to maintain your own routing/switch/etc infrastructure.
- How you see a dedicated vendor providing managed leased server hosting, network services, on-hands management, etc, as to be genuinely different from a cloud provider -- other than requiring a significantly longer turn-around on provisioning and management tasks.
- How expensive (and for what length) the lease terms are on that server hardware. I've yet to find a quality managed hosting provider that will lease hardware at terms that come close to matching the pricing of either in-house maintained or EC2 provisioned servers).
- probably they are not very different than a cloud provider in that aspect, other than that I seem to be getting a pretty good treatment from hosting providers in general (EV1/The Planet excepted, they've gone downhill to the point that we quit hosting there).
There are differences between providers, but for the most part those are relatively small once you reach a certain level.
- The lease terms are variable depending on the use case, but the majority of the longer term leases are for a year to two years, flat rate published price is about 1E29 / Mbit exclusive VAT for 100Mbit, including server lease. I get a better deal than that but I've been asked not to publish it, I'm sure you understand.
A box with 24 1T drives, 8G of ram and dedicated 1G flat-rate uplink currently lists at E1199/month ex vat if you pay in one go for a year, a bit more if you pay month-by-month.
If you're located outside the EU then you do not pay VAT (and if you are in the EU you'll get it back).
Good negotiators will probably be able to shave some off that price, and if you are able to serve lots of bandwidth with relatively little cpu you can add another G for E750, which puts you under 1 euro / Mbit.
If I would do the same using Amazon I'd be paying a multiple.
They're not "manag[ing] it themselves" -- it's a gigantic stretch to say that software RAID on an instantly-configured EBS volume is remotely equivalent to specifying, ordering, building, installing and maintaining a number of RAID arrays in a data center.
Basically this seems to put a fairly low upper limit to using 'the cloud' for something a little larger before you get back exactly the same kinds of issues that you were dealing with when using self-hosted hardware, only at a higher price point.
This is ridiculous. The primary issues with self-hosted hardware:
- Energy usage. Data centers only have so much power and cooling to spare, what's available costs money, and you constantly have to monitor/juggle utilization and search for more energy-efficient hardware, which leads to the next issue ...
- Physical plant. You often need to buy it well ahead of time to ensure that you'll have power and space available for expansion should you need it. You need to wire that space, buy switches and routers, and that leads to the next issue ...
- People. You'll need people to order, build, and install those servers. You'll need people to swap out dead components in those servers. You'll need trained people to install PDUs and console management systems, to configure and run your switches and routers.
EC2 and similar pushes these issues upstream. You can manage your servers in software. You don't have any routers or switches to manage, you don't have to order, build, or install hardware. You don't have to expend significant capital outlay on servers or rack space.