DeskRoll consists of the web part, the communications server and the client applications. We offer deep the customization of all three. We can brand the website and the client apps and we can deploy the communications server where you deem appropriate.
Why You Would Want a Dedicated Communications Server
The main concern is of course security. With a dedicated communications server, you will have the client applications that only work with your server. That means only your technicians can access the computers with your custom apps installed.
The second concern is speed. You might have an environment that does not allow peer-to-peer connections. Deploying a dedicated server should increase the speed and reduce latencies.
What We Can Do
Having broad expertise in the remote management of infrastructure, we can remotely deploy in virtually any environment of your choice. Obviously, there are a lot of options for you. If you are interested, contact us and we will advise.
Also, let us consider them in more depth. We will start with the basic concepts and notions.
Concepts Behind Cloud Computing
Utility computing concept is the basic concept behind virtualization, cloud computing and the bloom of on demand computing and thin clients. Utility computing approach enables metering of computer resources which is necessary for their efficient management and rental. This allows to allocate the appropriate amount of resource for proper performance.
Grid Computing and Distributed Computing, Virtualization
All three approaches allow you some flexibility in allocating and sharing the resources among different tasks.
Both grid computing and distributed computing are aimed at doing one task on multiple physical hardware units (nodes). They nodes might be even similar or identical in terms of hardware. Also, both approaches assume that the nodes accept multiple data for input and then they produce the end result.
Grid computing is similar to distributed computing. Still, the latter envisages a more dispersed system where the end result is combined from the data output by all the nodes. As for grid computing, it sees the entire array of nodes as a homogeneous pool of resources. This way (unlike distributed computing), it provides for the virtualization of resources.
Virtualization enables more efficient and granular management of computational resources, exposing standard resource units for rental. It makes cloud computing a viable concept and business model.
Virtualization abstracts from the actual hardware and exposes some pre-defined "standard" resource units instead. Normally, pricing and licenses is based over those resource units. Some resource units can be fairly tightly associated with the underlying hardware, like the amount of RAM. Some might be less, like rented abstract consumer-class "CPUs" or "CPU cores" (the underlying hardware may have nothing in common with the actual consumer-class units).
Cloud vs. Self-hosted
Costs
To host something on the premises you need to have space, infrastructure and human resources.
This makes self-hosted solutions rather expensive. The costs may largely depend on the software platforms, however. Some software like Windows Server can be quite expensive. Depending on the setup, the costs might be tens of times higher than the hardware/infrastructure costs.
As for human resources, outsourcing infrastructure management and maintenance might reduce costs. But normally you would have in-house experts to manage your in-house infrastructure 24/7. Still, renting human resources might be a viable idea, once you've considered the possible risks and implications like security, uptime, etc.
Data centers have more economy of scale, so the infrastructure, hardware, software and human resources come cheaper for them. So, colocation can be considered as just a deeper kind of outsourcing where you rent infrastructure but you may have you hardware and software. Colocation is the choice for serious projects with tight and/or very specific hardware requirements where you want high-performance dedicated hardware. Otherwise, you might go for simpler options like managed hosting (a dedicated server).
Of course with classic hosting, you know how much computing power and space you need. So, you rent it pretty much the same as with the cloud services. So, what's the difference?
Cloud is flexible and scalable. The essential concept of cloud computing is the dynamic management of resources, enabling you to cope with sudden peak loads. That's what makes it different from virtual private servers. Normally, with a VPS you only have disk space and RAM guaranteed, while CPU is not. These fundamental differences are normally reflected in the service level agreements.
Reliability
Reliability is provided through resource redundancy and geographical distribution.
For on the premise hosting, deploying failover hardware (and software) multiplies the costs. Geographical distribution may result in a massive cost increase that only large companies can afford for their specific tasks.
This way cloud hosting may provide more reliability at a lower cost, depending on the provider.
Security
If you are hosting on the premises, it's generally easier for the intruders to find out the exact location of your data. Then it's only the matter of resources that are needed to break and enter or simply destroy the hardware.
Having your data and applications at data centers is where security normally comes at a lower cost.
Again, this is basically the matter of resources that you invest into security.
Deployment and Configuration Times
On the premise hosting and colocation are obviously the options that are the slowest to manage. To make quick changes to the configuration (including emergency hardware replacement), you need to have hardware parts in stock. Otherwise, shipping might take weeks.
Again, cloud has an upper hand here again, with its ability to manage resources on the fly.
Conclusion
It might seem that cloud hosting is the choice for about any project out there. For many projects, it is.
However, there are many aspects to be considered before making a decision. And the main driver of your choice is, of course, the budget and resources at hand.