ISPs oversee their networks from a Network Operations Centre, which usually falls under the Technology department. This Operations Centre is essentially a large control room featuring numerous screens that display various information such as network usage, providing warnings if specific metrics fall outside the acceptable range.
ISPs tend to concentrate their monitoring on the core network as it is the primary conduit for all data and any issues here would affect a significant number of their customers. There are several approaches that ISPs might follow to monitor their networks, which for simplicity's sake, we can categorize into three main types:
Reactive Monitoring: Some ISPs adopt a basic, reactive model, where issues are primarily identified through customer complaints. The resolution method involves dispatching technicians to customer locations.
Network Management System (NMS) Driven: A more advanced approach involves the use of a Network Management System, often provided by a major equipment manufacturer (e.g., Nokia, Huawei). These systems provide data on system monitoring metrics, such as CPU usage and bandwidth utilization. However, this measurement approach isn't explicitly issue-oriented, and not all problems can be detected using this method.
Advanced Data Monitoring and Automated Action: The most progressive ISPs might take one or both of the following steps, though a limited number of ISPs globally have achieved this level of development:
Issue Identification: Implementing a system that focuses on capturing data that identifies a broader range of issues than can be detected by an NMS alone. This system is usually provided by a third party.
Action: They have set up automated actions in response to known issues. For example, they might reroute traffic or activate additional resources once capacity reaches 40%
The team in charge of network monitoring at the Network Operations Centre is usually led by a mid-level management employee. Their reports to senior management typically focus on average network utilization over the previous period. However, this approach doesn't emphasize issue identification or resolution performance and may not be leveraged to its full potential within the organization.
Last updated
Was this helpful?