Hyve Cloud Hosting vs. Basic Cloud & VPS
This page describes the key benefits of using Hyve Enterprise Cloud Hosting and illustrates why our premium Cloud Hosting platform is superior to the basic Cloud Hosting (or VPS) offered by many other hosting providers.
VPS & Basic Cloud Hosting
The ability to partition a physical server into multiple 'virtual servers' or VPS (Virtual Private Servers) has been widely available for many years. The process involves taking a standard rack-mounted physical dedicated server and using software to create multiple instances of the operating system (OS), making it possible to provide multiple 'virtual' servers from a single 'physical' server. This approach offers many specific advantages; primarily a reduction in hosting fees.
Since the concept of Cloud Hosting has become fashionable, many hosting providers are promoting basic VPS as Cloud. The reality is that the two are very different; the key difference being that VPS involves one physical server spit into multiple 'virtual' servers, whereas true Cloud Hosting uses custom designed hardware comprising multiple physical servers, hard disks and controllers to offer genuine benefits.
The problem with VPS & Basic Cloud Hosting
The main problems with basic Cloud Hosting & VPS are poor performance, reliability and scalability. By splitting a single physical dedicated server into multiple 'virtual servers', a huge strain is placed on the various components of the server.
Even the most powerful traditional physical servers were not designed to cope with this type of load. The processor(s) and memory that were once being used for just one server are now working for multiple servers. The same is true of the hard disks; the quantity of read/writes being requested increases dramatically and in many cases, the disks cannot keep up with demand, a huge bottleneck is created and the website being served grinds to a halt.
Hyve Enterprise Cloud Hosting
Our new Enterprise Cloud Hosting platform addresses the problems associated with VPS by using the industry's premier custom-designed hardware and software.
We have invested in purpose-built HP hardware comprising powerful Blade servers and Fibre Channel disks which have been specifically designed by HP, the industry's best hardware manufacturer, to address the problems associated with server virtualisation.
We have combined our careful selection of hardware with the use of VMware software to control resources. This combination results in a hosting platform which provides unbeatable levels of scalability, reliability and performance.
High Performance
- Brand-new Blade servers featuring state-of-the-art Xeon CPUs & very fast Memory;
- Minimum of 16x Fibre Channel (FC) disks working in parallel serving data. 16x FC disks instead of 2x SCSI disks as often found in basic Cloud Hosting;
- Disks connected via multiple Fibre Optic cables rather than a single Serial cable;
- Server to server communication is much faster. Blades interconnected via Backplane technology, not Ethernet;
- Internet uplinks operating at minimum of 1Gbe and fully burstable to maximum capacity.
Reliable & Fully Redundant
- Every Blade server has at least 4x Fibre Optic links to the disk array via 2x independent fibre controllers (a redundant pair). A single serial interconnect is the norm on stand-alone SAS drive servers;
- 2x dedicated controllers per disk array instead of a single SCSI controller on stand-alone servers (which also adds load to the server CPU as well);
- Every Blade server benefits from 6x separate power supplies per enclosure rather than the single PSU in a stand-alone server;
- Every Blade server benefits from multiple enclosure fans. If one fan fails, the others increase RPM to offset the loss. Again a single point of failure in a stand-alone server;
- The Blade enclosure has 4x 1GbE internet uplinks for redundancy and capacity, not 1 as found in stand-alone servers;
- The Blade servers operate in HA mode (High Availability), so servers can instantly migrate between Blades in the event of a hardware failure. A stand-alone server would simply fail;
- Multiple Cisco firewalls are in place for security. All operate in HA mode, so if one fails another will take over instantly;
Scalable
- Rapid scalability; we can upgrade hardware resources with a single reboot of the guest OS;
- Blade servers have LS technology (Load Sharing). This enables each Blade to utilise another Blade's CPU resources if permitted. This is purely for traffic peaks and spikes and is only enabled if requested and will not affect other Blades because it will only utilise resources from those currently underutilised.
