Re: What can I use a Cloud for
A cloud is an abstraction.
The idea is that you have a "farm" of VM hosts, either VMware or KVM/Qemu, or whatever. These hosts are all pretty much the same, and the cloud manager tries to keep them balanced for load by some strategy.
Now you, or the IT staff of the company, get need for a server. You requisition one from "the cloud" with specs. In a few minutes, there's a VM and it has some sort of operating system on it per your specs.
That server resides on the cloud, but it's no longer an abstract server. It's configured the way you configure it just like a real box. You don't know exactly where it is, the cloud manager takes care of that.
The only things you really need to know about a cloud server:
- What operating system is on it.
- How much RAM
- Hard disk space (and possibly speed)
- Firewall settings
- What software is on it.
- How heavily it's used.
Now, what you use it for: It can be anything, but you need a client or clients to get that information. You could have a web server, or an application server, or a mail server, or whatever. You could have some sort of computation application going on it (weather modeling or something) or anything else you can dream up.
The questions you're asking on this thread, it seems you don't see the point for a server, much less a cloud. That's OK, generally speaking for personal use there isn't one.
A company owner or IT manager would want a cloud so they can:
- Make full use of hardware they have.
- Measure how much of their hardware is used.
- Have quantifiable space and capability for expansion
- Handle system management in an abstract and consistent way.
- Have failover in case one host goes down.
- Anticipate need for hardware upgrades
- Enable hardware upgrades without loss of service for users.
Some virtualization schemes can move a server from one host to another without shutting it off.
Now for a server:
You've mentioned a server vs having an office app on each computer. Some reasons for the centralized approach:
- If you have software installed on every workstation:
- You have to make sure the software is installed and working on every computer.
- If a disk fails you lost a license, and probably the documents somebody saved on that disk.
- You can't really control what each employee sees.
- Software version control is problematic.
- Your IT staff bounces around from computer to computer to get their work done, at which time the employee who usually sits there is paid to stand there.
- If you use a server (e.g. Microsoft Terminal Services or an equivalent X server setup for Linux):
- Everyone has the same version of every app.
- Their documents all get stored in a specific place
- Your IT staff can back up that place every night with a centralized backup.
- The clients can be pretty much anything, and don't even need a hard disk.
- System changes are done by the IT staff without having to interact much with users.
- A server on a public cloud:
- You no longer have to support hardware maintenance.
- You pay a monthly fee for the server, which is a predictable cost.
- Somebody else handles backups
- Your employees can work at home or at the office, pretty much the same thing.
- Upgrading a server means you specify the upgrade parameters, schedule a time and the server is upgraded.
- Duplicating a server means you provision it, and use the first server as a model. Again, a few minutes and it can be done, depending on how much data is involved.
Now, the difference between a public cloud and a private cloud is that with a private cloud you set up your VM hosts as you want, and nobody has access to the hardware except you and your company. You perform the maintenance, you buy the hardware, your people are on the hot seat if something breaks. A public cloud, somebody else does all that and there's an extremely good chance that they'll have your information up and running WAY before your personal guys could do it.
Help stamp out MBR partition tables. Use GPT instead!