Before the introduction of the Cloud Computing systems there are two ways to deliver web hosting services - using stand-alone physical or virtual servers or some form of cluster pf physical computers (servers). The most common one was to use stand-alone physical servers to host websites as Virtual Hosting accounts. Virtual Hosting (a.k.a. Shared Hosting) is a method for hosting multiple websites and domain names on a stand-alone server. The website accounts are divided into different directories and share the as CPU, memory and all the resources on the underlying physical server.
Most Virtual Hosting (Shared Hosting) accounts are "Name-based". "Name-based" means that there are multiple host names and/or domain names running on the IP address used by the physical host server. For example the server can receive requests for domains, domain1.com, domain2.com, domain3.com, domain4.com, domain5.com and many more. All of them resolve to the same IP address. However when the server receives for example a HTTP or a HTTPS request for domain1.com, it sends it to a HTML file placed at directory /var/www/user/domain1/site/ and so on.
The downside of the Virtual Hosting has always been the server load, which different accounts create when they need to use a lot of resources. Before Cloud commuting there was only one way to overcome the server load issues. It was to create some form of a cluster of physical servers, which divide different computing tasks between them. Although technically possible, this wasn't a very common method of improving the Virtual Hosting systems, due to the way that server automation software (control panels) used to work. They usually lock to the CPU and physical hardware on one of the physical hosts and thus were unable to function properly in a multi-server environment. So, before the Cloud computing, which could be considered as a computer virtualization that works in a multi-server environment, those who used to do Shared Hosting, didn't have much choice. They had to deal with server loads on a daily basis without being able to distribute the load to another instance or to scale the CPU and RAM resources of the physical or virtual server that hosts the Virtual accounts.
How Cloud computing changed that?
Cloud computing has improved significantly the Web Hosting in each service niche. In the VPS niche for example the Cloud Infrastructure allows the isolated Virtual Servers to be scaled up and down instantly. In the Virtual Hosting (Shared hosting) niche the Cloud systems allow not just any virtual Cloud-based host to be scale up. They allow High Availability and Failover.
High Availability (HA) is a technique which reduces any possible outage on a computing system by making possible for this system to return back online very fast after a failure of its Operating System (OS) or a downtime on the physical server, where certain virtual host used for delivery of Shared Hosting services resides. HA is usually a function that monitors the computing instance for overload, OS failure or a downtime. If any occurs it simply restarts on the same physical server or if it is down - and here comes the Cloud computing - on a different physical host, part of the Cloud computing system. The main purpose of High Availability is to reduce the service downtime, not to prevent it. It simply brings back the virtual server used for delivery of any computing service online very fast.
Failover refers to the ability of a Cloud computing system to continue delivering any services without an interruption, in case of an OS or a hardware failure. There are different failover technologies and scenarios. What is important is that Failover (or Fault-Tolerance) is a function that in case of an outage, transfers the workloads from any failed computing node (server) to new one.
Before Cloud computing, web hosting service providers had to deal with server loads, suspend the virtual account which used to overload the underlying server, had to migrate them on a new host or simply to put them in an isolated environment - on a VPS or a dedicated server. Nowadays, the Cloud computing systems allow the providers to scale up in real time the computing resources of the servers they used to deliver web hosting services. It is a Cloud Web Hosting service! When High Availability is available as function, it reduces the service outage to less than 1 minute, while the failover simply prevents service downtime by either duplicating and mirroring any cloud server instances or by automating the real time service migration, when a failure is detected.
We came to the conclusion. It is "Ask the web hosting provider you want to go with do they have a Cloud infrastructure in place and does it provide High Availability and/or any form of Failover to the Web Hosting accounts".
