One of cloud computing’s chief definitional benefits, as Citrix contributor Tom McCafferty notes, is the ability to not simply hire a company to create a cloud for you but to actively pursue self-service options with a range of different providers.
However, there are things to consider for choosing either managed services or self-service pathways to the cloud. While we have covered at length the virtues of managed services, we thought it would be worthwhile to compare the benefits and drawbacks of self-service.
1) Speed: Self-service is faster than opening a ticket, which is a great ability especially if you’re a time sensitive business.
2) Onboarding: If you have a customer profile and you’re bringing up a new customer, you can deploy infrastructure more quickly by pushing a few buttons. Root access: Let’s say you have a self-service cloud and you spin up a bunch of servers part of that process would be getting root access to the servers, IP addresses and admin access etc.
3) Application requirements: If you have a complex application and you want to measure latency or higher level networking or load balancing rules, self-service combined with a measure of root access will enable you to configure VMs to match the needs of your application.
1) It’s on you: Self-service means that all parts of infrastructure, with a special emphasis on compliance and security, but also including firewalls, switches, load balancing rules, ensuring HA and backups have to be handled internally, which also requires such expertise to exist in-house.
2) Customer provisioning: Self-service not only means procuring infrastructure for your business, but also setting up the infrastructure that will support your clients, including their own networks their own security rules, DMZs and all services. If you do not have the staff for this, it may become detractive to your ability to innovate around your company’s core value, since much upkeep is necessary for infrastructure builds.
3) Root access: In some cases, root access can be a negative, because it is precisely so operational. Companies also may not want their staff to have to install things on the servers themselves. They just want then to spin up the servers with limited access (to a directory to write code for example) but not root access to servers.
4) Application requirements: There is a trade-off with self-service – you want to have access to what you need but you don’t want to mess things up. Well written applications don’t need as much support, or if the service provider is giving good latency andset up – network wise, they shouldn’t need self-service. If you’re writing things on the fly and having problems, having access can be requested.
In the end, self-service and management are not polar opposites. In an ideal situation a managed service provider can work with your business up front to build the environment and then embed self-service capabilities consistent with security and compliance objectives –network, security, firewalls, scanning VPN, and encryption – that you can access to spin up resources for your clients. Balancing time and resources should be the ultimate measure of whether self-service only or managed services only is the optimal process of infrastructure management your business will benefit most from.
How does your business determine whether self-service or management is the best path? Let us know on Twitter @CloudGathering.
By Jake Gardner