In our last two posts, we’ve discussed the basic definitions of cloud services and a few considerations when moving to the cloud. In our final post of our three-part series, we would like to discuss practical configuration examples we would use when utilizing “the cloud” for your infrastructure. (We know, these posts are a little techie/geek-speak… We’ll come back down to earth in our next post… we promise. If you prefer something a little more “techie light”, we suggest THIS LINK!)
In the scenarios presented, we assume a company will require a Domain Controller (IE: your “network brains”, remote access (aka Terminal Server), and email (via Exchange.)
PRIVATE, ON PREMISE “CLOUD”
A “host” machine would be procured and built utilizing Virtualization technology such as VMWare or Hyper-V. This system would become the host platform to install a Domain Controller/File Server, Terminal Server, an Exchange Server. A new Backup Appliance would be implemented (Barracuda Network’s Backup Appliance – which also has the option to store data off-site)
Pros: The inherent benefit of a Private cloud would be the control and flexibility of server ownership. This option would also most-likely be the most cost effective over the long run – despite the need to refresh servers moving forward.
Cons: Server closets will need cooling, they may not be accessible when needed, and there is likely not redundant power and internet to ensure the system is available all the time.
PRIVATE, DATA CENTER “CLOUD”
Your company’s infrastructure would be identical to the “On Premise” solution, however rack space in a data center would be obtained to house physical infrastructure.
Pros: Having Data Center rack space provides redundant power, fail-over internet, and climate control. Typically Data-center personnel are on-call to help reboot servers as needed. Security is also provided.
Cons: Accessibility to the Data Center is somewhat limited as any users entering the data center (including third-party support personnel) will need to go through authorization processes. Any user accessing these servers/services will require Internet access.
Servers would be a fully redundant solution with a dedicated Virtual Machine instance for each of COMPANY’S servers (Domain Controller/File Server, Terminal Server, Exchange Server). The systems are all dedicated virtual machines on shared hardware infrastructure. The hardware is clustered and runs on a fully-meshed network. The infrastructure is architected for 99.995% uptime.
Pros: There would be no ongoing need for capital hardware purchase for server infrastructure. Having multiple Internet connections at the hosting provider and redundant hardware would provide 99.9% uptime. Office 365, if utilized for Exchange, is a fully vetted product and would provide SharePoint services for document management and work-flow.
Cons: If only one internet connection is available at a company’s office, this solution has single-point-of-failure at that connection. The loss of internet service which would result in downtime for the entire office environment. Cost is also a consideration as Public Cloud computing is still considered a premium service and is typically more expensive (over on-premise infrastructure) within 3-4 years. As a company’s data footprint increases; future Cost per Gigabyte for storage needs to be considered as an additional/added expense. Backup retention is currently 14 days. If longer retention periods are required or desired, an additional monthly fee would be incurred and will be based on the data size and retention period. Hardware upgrades would still be required if the Hosting Provider forced an upgrade to server software which would require an upgrade to client machines/software, etc.
Whew, we’re done. The bottom line… there are lots of things to consider when designing your computer network… We’re happy to do the heavy lifting for you… just give us a call – we’re here to help!