- Posted by Ramya
- On January 4, 2021
- 0 Comments
An EMS periodically collects various performance metrics from each network element (NE) in the network to accomplish the following:
- Monitor the performance of Network Elements (NEs)
- Enforce the Quality of Service (QoS) (based on the service level agreement)
- Raise alarms whenever there is a deviation from the target QoS
- Provide real-time views or reports of the NE’s performance
- Upon request, provide the user with historical graphs or reports of the NE’s performance
Challenges to overcome:
While an EMS is trying to perform such tasks, it faces multiple challenges such as:
- The task of collecting performance metrics from the device is easier said than done. The efficiency of the EMS depends on various factors including:
- The number of metrics that need to be collected at a time
- The number of NEs to be monitored (which increases the load on the EMS)
- How frequently data is collected (this depends on the importance of the metric)
- The network traffic due to the EMS should be kept to a minimum so that activity monitoring does not degrade the performance of the NE and does not cause congestion in the network.
- The bigger issue here is the efficiency of the EMS. The efficiency becomes noticeable when the EMS is used in larger networks. To a large extent, the efficiency of the data collection engine depends on its design. Therefore, care should be taken to ensure that the performance collection methodology is well designed.
Here are some design strategies that can help improve the efficiency of the data collection engine:
- Since data collection is mostly an I/O operation, the CPU will not be busy 100% of the time. Data collection can be performed in a multithreaded environment by collecting data from multiple devices in parallel. This gives way to support more devices. We recommend using a thread pool component for data collection in order to control the maximum number of concurrent data-collection activities that are being performed. This provides control over scarce system resources.
- It makes sense to prioritize data-collection metrics into various groups and schedule them in different data-collection intervals. This allows for frequent collection of the most important performance metrics while less important metrics are collected at less frequent intervals.
- Another key strategy is to make use of intelligent data-collection components that have the ability to group related metrics and collect data using Get-Bulk requests. This reduces data-collection time and lowers network traffic – a critical requirement for an EMS. The scalability of the performance module is largely determined by the design quality of the data collection component – which in turn determines the scalability of the EMS.
- A well-designed data collection engine also reduces the overall load on the EMS by performing bulk updates to the database (another expensive I/O operation) instead of performing individual inserts.
Utilize the aforementioned tips to formulate an efficient EMS. Or, you could trust the job with the best EMS platform – Dhyan’s NetMan has been managing over a million devices for more than 16 years. NetMan comes incorporated with the above design strategies, helping you to seamlessly monitor and control your devices. You could also make use of the varied customization services Dhyan offers for your EMS.