MobiledgeX Edge-Cloud 3.0: Improved Visibility and Operational Management of Cloudlets for Operators

August 25th, 2021

Headshot for Wonho Park

Wonho Park

Chief Product Officer & SVP Product Development

In last week’s post entitled “MobiledgeX Edge-Cloud 3.0: Optimal Edge Connectivity with EdgeEvents” (by Vasanth Mohan) we discussed the improved edge connectivity via MobiledgeX EdgeEvents for developers. In this post we’ll be discussing the improvements we’ve made to the visibility and operational management of cloudlets for telecom operators.  

MobiledgeX Resource Management

With MobiledgeX Edge-Cloud 3.0 operators can now start leveraging our controller as the resource manager rather than relying only on their IaaS layer to determine the resource availability of cloudlets. When creating cloudlets, resource quota limits are defined such that the MobiledgeX controller can track resource usage. The advantages of this feature include receiving notifications through our alerting system when resources are impacted and have exceeded the defined threshold, which allows operators to respond quickly by modifying or increasing the resource limits and capacity. Figure 1 shows an example of a resource quota to manage 712 GB of RAM, 4 GPUs, and 180 vCPUs and to generate alerts when the Alerts Threshold reaches 80% of the respective values. 


Figure 1. Setting Resource Quota 

 

Cloudlet Alerts with Alert Receivers

Starting with our MobiledgeX Edge-Cloud 3.0 release, operators can begin to monitor the performance and counteract irregularities within the system using our alert framework, which helps mitigate any performance or functional issues. Alerts generate when the following occurs: 

  • Cloudlet is offline

  • Resource limitation has exceeded

A notification is sent either through Slack, PagerDuty, or email, depending on the preferred delivery method configured by the user. Once the issue or condition is resolved, an additional notification is sent to the user. Figure 2 shows an example of an email sent to the organization's manager when vCPU usage exceeded the threshold.

 Figure 2. Sample Notification Email

Cloudlet Usage Reports and Scheduler

With the controller now managing resources, the platform can store historical data of the cloudlet resource, allowing operators to directly retrieve cloudlet usage information through generated reports detailing the usage. The MobiledgeX Edge-Cloud 3.0 platform can determine the time zone and provides a configurable calendar to run the operator reports based on the interval specified by the user. Additionally, periodic scheduled reports are generated and sent to the user based on defined intervals by the user. The figure below shows an example of a report scheduler configured to auto-generate reports at the intervals of “Every Week” and “Every 15 days”.


 Figure 3. Configuration of Scheduled Report using MobiledgeX Console

A sample operator report is also shown below. The report information may include, but is not limited to, the following information:

  • List of cloudlets and their platform type

  • Cloudlet's last known state

  • Flavor usage count by various applications

  • Cloudlet alerts

  • Application state

  • RAM and VCPU usages

Figure 4. Snippet of a Sample Operator Report

This delivery mechanism provides operators with the data they need, when they need it, to understand how their cloudlet is being consumed and running. Furthermore, the information is used to provision their cloudlet resources judiciously.

Using Kafka to push cloudlet events

Starting with the MobiledgeX 3.0 release, operators can manage real-time data streams using Kafka. By providing Operators with events streaming API, events can be pushed directly to a server (Kafka cluster) that the operator provides to MobiledgeX. Currently, MobiledgeX supports pushing cloudlet-related events to the Kafka cluster. 

Monitoring: Event, Usage, and Metrics

Providing monitoring capabilities is critical to ensure the performance of operator services. As an operator, we understand that having insights into operational events and metrics can help operators mitigate issues quickly, thereby reducing downtime for developers who have applications and cluster instances deployed to their cloudlets. We made significant improvements in this release with new search capabilities and additional events. Our new usage logs also help operators understand the application activity occurring within cloudlets. The below figure shows our improved Monitoring Dashboard to help operators visually centralize, collect, aggregate, and analyze events, logs, and traces so that they can get a bigger picture of what is going on across your infrastructure in real-time. 


Figure 5. Screenshot of MobiledgeX Monitoring Dashboard for Operator

Conclusion 

The MobiledgeX Edge-Cloud platform continues to provide operators opportunities to take control and manage cloudlets and their infrastructure through our easy-to-use console. Accessing critical information and resources -- and gaining insights on cloudlet resources and usage -- is an invaluable tool that operators can continue to leverage to monetize their edge clouds.

At MobiledgeX, we are inspired to look for ways to improve the operator’s experience. Operators continue and benefit from using their own telco network infrastructure alongside our MobiledgeX Platform to achieve a consolidated workflow with minimal overhead to the operator’s bottom line.

To learn more, visit the MobiledgeX Operator Portal