How to coordinate the cloud management between provider and enterprise

Effective management of application performance and availability in the cloud seems to be a matter of visualization and consolidation, but it actually stems from the relationship between application components and services. To solve this problem, you must first know how the end-user views the problem.

The end user who sees the cloud as a simple tool in the perspective of server consolidation treats cloud management as a way to secure and activate cloud services. Cloud management is a key factor in ensuring a quality experience (QoE) for users who view the cloud as a resource in the Extended Application component network.

Not surprisingly, the latter group is committed to the cloud application management strategy.

QoE is the sum of quality of service (QoS) in the Workflow network connection, plus the application component flow of the network link. In the cloud, components are allocated to resources at different locations in a resource pool. Obviously, if the application moves from one cloud to another or from one provider's data center to another, network connectivity changes, but other factors can affect network performance, such as storage area network congestion or a larger load on a server hosting a virtual machine (VM) in a new datacenter.

These performance issues are not directly visible to enterprise management tools, since they all occur inside the cloud, so any information about these public resources comes from the cloud provider.

One way to integrate cloud management data and corporate networks and it information is to use the cloud operator's management level as a service (Mpaas). Essentially, Mpaas creates a management window for a cloud provider's management system. The Integrated management view of the cloud provider links a corporate management API into a unified view. This provides optimal management visibility and ensures that enterprise users and their providers have a single view of the cloud resources, facilitating service level agreements (SLAs).

Only a handful of cloud operators, such as HP, IBM and Microsoft, provide integrated management services, and most are designed to do so on specific orders, but industry insiders believe the future will do more. For now, the expected platform is the service ditto or large it provider with more functionality.

You may also create an integrated management view, even where Mpaas is difficult to implement. The problem is how much cloud management data can be made available to cloud providers and how do they increase on demand?

Some cloud providers will provide management statistics for professional examples. The use of professional examples, such as managed applications constrained by geography, will help stabilize network variables and may also reduce server VM performance variables.

When statistical professional instances are not available, you can install some management components as part of the cloud hosting application image. At a minimum, this can provide an application view of local resources and connectivity performance. Before installing any Third-party tools, be sure to check the cloud provider's management tool compatibility.

If the Mpaas fails and the visibility of the server and network connectivity within the cloud is not available, the management responsibility must be based on the participants, and the contact points must be established to allow SLA writing and enforcement. The goal is to reduce the number of areas included in management and improve network visibility. The network connects end users to the cloud, creating a potential management black hole. Try to make the cloud provider responsible for network connectivity or use virtual private network services with specific monitoring and management. Helps reduce the impact of visual network performance changes and simplifies SLA execution.

If all of these are invalidated, the management boundary applies the connection point participant settings contained in the public cloud or mixed cloud. The main variables are networks, especially when the Internet is involved.

It is easier to find field trends or changes by starting a measurement that applies response time and then subtracting the measurement latency, and pointing out the QoE problem. After that, you can directly ask an entity to take responsibility for the cloud infrastructure. This method is better than nothing, but most users still crave Mpaas.

(Responsible editor: Fumingli)