The rise of cloud computing had also lead to the rise of cloud management systems. Cloud computing, at its essence, is a complex distributed system with many moving parts, which makes these systems a requirement. To keep those parts all working and playing well together, you have to have software that’s able monitor, govern, correct, and control.
Cloud management provides the ability to operate and monitor single or multiple cloud instance(s), including IaaS, SaaS, and PaaS, as well as underlying server or application instances. It’s about keeping things working and under control while leveraging cloud-based resources, as well as monitoring core subsystems.
The types of available cloud management systems vary greatly in features and functions, even in the method of delivery. Some are on-premise and provide cloud management from the inside out, monitoring public and private cloud computing services. Some are hosted in public clouds, providing monitoring and management services on-demand, typically only monitoring and managing public cloud services. Some come from more traditional technology, such as performance management systems that are retooled for the clouds, or re-purposed services governance systems from the world of SOA.
Within the current maturation of the industry, no two management and monitoring systems are alike. Thus, you may end up with one, two, or even three management and monitoring solutions to reach your cloud management objectives.
When assessing cloud management you should consider the following capabilities:
- Availability monitoring, including alerts for system and/or component outages.
- Cost monitoring, including alerts for components that cost more than expected.
- Performance monitoring, including monitoring performance and adding or reducing the number of resources on-line.
- Service monitoring, the governance of all APIs (Web services, typically), either for consumption by external systems, or APIs that are being consumed by the system in our cloud.
Availability monitoring is provided by cloud management solution providers or by the cloud provider itself. Typically, this alerts the cloud manager when the cloud service is not responding due to an outage or communication issues. Thus, the cloud manager has an opportunity to correct the problem,
In some instances, the cloud management system can take automatic corrective actions, such as restarting a system, or de-allocating and reallocating a problem resource, such as a virtual server within a public or private cloud.
Cost monitoring becomes a critical part of the cloud management system in that storage, compute, or databases that get out of control could end up costing thousands more dollars than expected. Typically, this portion of the cloud management system allows you to set thresholds that will generate alerts around excessive usage. Also, in some instances, they will allow the management system to take automatic corrected actions.
Performance monitoring has two purposes: First, the ability to monitor the performance of the public or private cloud providers in real-time, either at the holistic system level or down to a core component. Second, is the ability to take corrective action, such as adding more virtual servers to serve the cloud-based system when the performance numbers begin to suffer. Or, perhaps removing virtual servers when they are not needed, thus reducing costs.
Service monitoring and governance allows the cloud manager to place a management layer between the APIs or Web services, and those that would consume those APIs or Web services. This allows the cloud manager to monitor those services for use, as well as place policies around those services which will automatically place limits on who can leverage the services, for what purpose, and to what degree.
Service governance systems are typically separate pieces of technology from core cloud management systems. However, their function when leveraging cloud-based systems is clearly mission-critical when considering the vulnerabilities and need for service/API management and control.
What’s key to cloud management is the ability to see all aspects of your system(s) that exist within public or private clouds, or sometimes both. As we continue to deploy systems within public and private clouds, we find that the migration or develop of those systems is rather easy when compared to operating those systems in the long term.
The problem is that cloud management is not a matter of purchasing or leveraging a cloud management system or service. To make this work, you must first spend the time to create a cloud management strategy that encompasses all of your cloud properties, now and into the future. Next, make sure you select and test the right technology. You’ll find that it will be 2 or 3 products and/or services that you’ll need to leverage, and the deployment of these systems will be complex work.
While this does add cost, in terms of managing your cloud-based systems, the use of this technology will insure your success in the clouds. Without it, cloud computing has no chance of providing the value required by the business.