WAN and cloud service management: Simplify your multi-vendor Environment

Most enterprises know the complexity of managing multiple service providers. These enterprises also realize that when they add cloud services to a variety of service combinations, especially when they use multiple cloud service providers, enterprises may face major risks of losing control, performance, and security. Fortunately, careful planning can reduce this risk to an acceptable level by carefully selecting suppliers and controlling the service deployment itself.

Cloud Services and WAN services share some common characteristics. Both have performance, security, and management issues. Both involve SLA, and may share responsibilities when they are mutually or together with other IT elements. We must solve these two problems at the same time, so we 'd better build a matrix of features and functions of cloud computing and WAN vendors.

Considerations for building your WAN/cloud service supplier Matrix

The first question for the WAN/cloud service provider combination given in the matrix is: whether the cloud service and the WAN provider provide a VPN connection, and if so, which supplier is the primary contractor. Enterprises should determine this problem for each cloud computing/WAN supplier combination, and they should also pay attention to the relationship between the two, because the longer their cooperation relationship, their support and diagnosis are as good as possible.

The second problem is their geographic location. Check the geographical location of each service in your map and determine whether this WAN/cloud computing combination supports this service location. If yes, whether the cloud service and WAN supplier have service points in the actual service area or need to return the service points. If the WAN supplier provides cloud support for a region from other places, a long connection may affect performance and reliability. This is usually the most complex analysis, because you need to analyze every major geographic location where your company wants to use cloud services.

The third problem is that it must cover monitoring and management. Does the WAN/cloud combination provide service monitoring at its connection points? The lack of monitoring at connection points means that the problem identification and fault isolation functions will become more difficult. Performance and status detection are particularly important in cloud computing networks because it is difficult to evaluate cloud performance if you do not understand the impact of network performance.

Consolidate SLA-build a sound WAN/Cloud Service Management

When a suitable combination is selected from the basic matrix, the next step is to review the SLA. When the network supplier is also a cloud service supplier, the enterprise may face the combined SLA and unified problem management. This means that the SLA can focus on traditional problems, such as normal running time and performance. However, if you have different networks and cloud service providers in any region, you need to focus your early SLA discussions on specific mechanisms agreed by both parties, such as performance standards and availability measurement. When both parties need to manage data, the way they analyze the data is very important. SLA cannot be based on subjective user-level performance and availability, but on the monitoring points agreed by both parties. Otherwise, you cannot confirm the problem and report the problem to the supplier. You also need a very specific problem reporting program, to determine how problems that cannot be solved by a front-line support contact are uploaded to a higher management level (supplier organization and the management level of your own company ). At the top of the reporting procedure, you should determine the arbitration process for the final settlement of the dispute.

When a supplier is selected and the SLA and contract are signed, the relationship will be transitioned to their operation phase. Most successful management of cloud computing users that contain multiple networks and cloud service providers said that their support process starts from the internal support of applications and their users, and then the team will filter the problem. In hybrid applications (networks and cloud computing), suspicious problems are usually first handed over to internal network support or IT support groups for their first-level problem isolation, then contact the supplier according to the requirements and SLA.

The Support Team of your own support program will need to contact the supplier. This will require special internal training and the development of an operational manual to describe how the issue should be isolated and reported. One problem that needs to be solved in these processes is how to "send alarms" when a problem occurs ". In many supplier cloud environments, problems in one place may occur on other management interfaces, and the notification information of this problem is spread to any visible place. If you do not, the same problem may occur, leading to expensive and destructive parallel resolution processes.

The last point is that the saying "people are everywhere" is not applicable in multi-vendor network/cloud configuration. As the number of suppliers involved increases, the complexity of these relationships also increases. The larger the number, the more you should specify a supplier as the integrator of the entire process. This will save a lot of trouble for identifying the problem afterwards, and improve the overall normal operation and user satisfaction.