Skip to content 
Introduction
In the Internet Service Provider (ISP) industry, key commodities are bandwidth and speed. Measuring and analyzing these two capabilities is critical to ensure that service-level agreements (SLAs) are met and that customers are satisfied. But with the growing complexity of networks, this is becoming a monumental challenge. Thus, it’s time to prioritize Service Assurance.
Manual and semi-manual assurance solutions are limited in their ability to benefit 4G, 5G, and cloud-based enterprise networks. The limitations include higher costs, more failures, and a longer time to diagnose and resolve faults.
Anuta Networks ATOM is the solution. It’s ‘active’ service assurance centralizes critical elements by aggregating all domains and network layers to provide an end-to-end service perspective. It is also a fully cloud-native platform supporting multi-vendor and multi-domain environments while enabling network operations to optimize IT and OSS functionality and reduce costs.
A Unified Approach to End-to-End Service Assurance
ISPs spend a vast amount of resources to probe network performance, diagnose issues, ensure quality, and monitor SLAs in order to satisfy customer requirements. Detection of service interruptions is often challenging due to a lack of mechanisms to sense service degradation and broken services. Network automation tools such as Anuta Networks ATOM are composite network automation solutions capable of detecting service degradation and monitoring, alerting, and taking remedial actions when required.
When it comes to testing a network, ATOM can check the provisioned services by two different means:
1) Through monitoring the OAM activities and operational commands passively and,
2) by utilizing a proactive approach in which ATOM sends the synthetic L2-L7 traffic actively to the network and evaluates the performance based on the behavior of this synthetic traffic stream.
The nature of the traffic stream can be defined and adjusted as needed and mimicked to real-world traffic. This process makes network services ready before commissioning and can also be used for active monitoring.
The ATOM Active Service Assurance (ASA) solution encompasses the following areas of network operations:
- Fault and event management investigation,
- Continuous performance measurement,
- Probe monitoring,
- Quality of Service (QoS) delivery,
- Service activation testing
- Network traffic calibration,
- KPI management.
ATOM ASA Architecture and Test Suite
The ATOM ASA test suite consists of test agents and test servers. The test controller is placed at a central location within the network as an on-premises, virtual, or bare metal server. It can also be deployed in the cloud as SaaS. The test controller is responsible for signaling the initiation and termination of tests conducted by agents. After test completion, the metrics are evaluated, and the result is displayed. Analytics and insights can also be generated from these test results with the flexibility to run multiple tests simultaneously.
Agents can be lightweight Docker containers, applications, or virtual appliances deployed on Linux machines and can generate the test traffic, which is routed through the desired path of the network. When the destination agent receives the test stream, the test metrics are evaluated by the controller, and the test result is published. ATOM ASA can also run these tests in monitoring mode, where the test agent continuously delivers the synthetic traffic stream for the desired amount of time in real time. It can also identify each test instance with a unique instance identifier, tracking the instance of the test which ran at a specific time.
Fig: Placement of ATOM Test Controller and Test Agent
Test Suite
ATOM ASA comes in handy to address critical networking requirements. It delivers a set of tests that measures network performance from Layer 1 to Layer 7, providing a comprehensive solution. As a vendor-neutral platform, it can also start and initiate tests on devices from any vendor for maximum flexibility. You can find some of the prominent tests below:
Pass-Fail Criteria, Error Seconds (ES), and Severe Error Seconds (SES):
Each of the tests has gauging mechanisms, where one can define a threshold for any test measurement or parameter, such as the occurrence of delay, loss, and jitter while TCP packets propagate over a link. If any service deteriorates below any threshold, a test is declared a failure. For every second of a predefined interval of time, if a service fails to meet the threshold, that is treated as one ES. Subsequently, if a test receives a delay and loss above the severe threshold level, then that second violation is treated as an SES. When a given service has violated the threshold, the color changes to orange, indicating a major service deterioration. Finally, when one receives severely errored seconds, it is shown in red, indicating that the service has been critically affected.
Service Unavailability
Whenever ten consecutive SESs occur, that portion of time can be treated as service unavailability. Also, the seconds not received in the form of a non-response from the client to a server are known as unavailable seconds.
Why ATOM ASA?
As networks grow more extensive and complex, there should be smart network analyzers to detect gaps in network performance as well as calibrate each functional aspect of the network. For instance, if a service provider provides voice services, they must check the network’s quality of Service (QoS) and how the UDP traffic behaves with different DSCP values, determine the traffic type that needs to be prioritized and allocate the appropriate bandwidth.
In ATOM ASA, such a scenario can be evaluated using a UDP test with a granular approach. A synthetic stream of UDP traffic can be injected from one side of the network by a test agent, which will act as a server probe with a specific DSCP value at suitable data rates, and the same can be received at the other end by the test agent (at the receiver end) as the client probe. Results can be obtained after the test, which provides insights into the actual performance of the network under suitable conditions.
Fig: Test traffic initiated by test agent from Location A
Fig: Test traffic received by the test agent from location B
Fig: Result metrics of the test
Fig: Based on the result, the SLA and availability of service are marked
Similarly, the performance of HTTP(s) protocol for a network can be measured by initiating HTTP(s) requests from the test agent acting as a client and based on the query and responses, can display the results of the test.
ATOM ASA Closed Loop Capability
If an SLA falls below expected levels, an integrated mechanism in ATOM ASA can create a trouble ticket by interacting with a ticketing tool over an interface such as REST API. The broader ATOM platform also has another feature called closed-loop automation which triggers a reactive action to any stimulus it receives. The alert can be defined based on the metrics the ASA test provides, and this can be used as a stimulus to activate a troubleshooting workflow or create a trouble ticket in any ticketing system, remediating the error quickly and easily.
Use Cases and Automation
ATOM ASA also allows network operators the ability to combine two or more test blocks to create a single test. For instance, if there is a need for a test suite that will test internet reachability and connectivity performance simultaneously, one can combine the HTTP, DNS, and speed tests to create a single test block. Similarly, to check the performance of site-to-site VPN, one can combine the BWPing test, TCP, UDP test, QOS test, ping, and traceroute tests to check the proper functioning of the VPN.
For example – let’s say an operator has to run the same tests between locations with the same parameters; then, he or she can schedule the tests as jobs to ease deployment. Whenever the test is scheduled, it will run and collect its result metrics. If required, the alert can be generated using CLA if the performance is compromised. Additionally, using ATOM workflows and APIs, one can integrate external messaging tools and email integration. Additionally, a test template can be created with suitable parameters and executed when required. The tests can also be triggered from external tools using API interaction, and a transaction can be initiated from an external tool to trigger the test.
Reporting and CSAT
Test reports can be downloaded in PDF or any documented format for flexibility. Based on the periodic tests and tests running in monitor mode, a precise SLA report can also be fetched. Subsequently, an ISP or enterprise can focus on what aspects of its services need improvement to enhance the user experience. Moreover, ATOM’s RBAC feature determines the control and access to the test results– such as who can execute the test and view the results and reports. ATOM ASA also offers a summary dashboard that provides insights into the network layouts of the services in an alert state. With multi-tenancy, users can be given control to execute the tests. This way, users can check the services, and an automated trouble ticket can be raised if the test fails for a considerable time. Since ATOM supports multi-tenancy and RBAC, users can have unlimited access to tests.
SAT Test in ATOM ASA
ATOM ASA can play an important role in service life cycle management when a service provider receives a new service order to provision an L3 VPN from a customer. The service provisioning can be triggered by a northbound API in the ordering system and sent to ATOM as the service order is confirmed. New services can also be provisioned based on network resource availability. Once the service is provisioned, ATOM can internally trigger ATOM ASA to initiate a service activation test. Once the newly provisioned service is tested and calibrated, the service can be declared as commissioned. Finally, whenever the change order is raised, the service ordering system can trigger the ATOM service orchestrator to make the desired changes in the network, followed by the service activation test to measure the performance of the change. If the provisioned service does not perform as per the KPIs, the user can raise a service ticket in any ticketing system. ATOM ASA is also compatible with TMF API standards to provide the ultimate degree of flexibility. To automate the service management lifecycle, ATOM also implements APIs such as the TMF API.
Future Plans
Dedicated private networks are expensive and take considerable time and expense to build. However, using existing assets with the appropriate performance assurance functionality unlocks the full potential of this promising approach. ATOM continues to lead this effort, with plans of incorporating AI in its portfolio to provide intelligence to test suites to enable automated route switchover and remediation.
However, today ATOM provides a complete solution capable of monitoring, alerting, and taking remedial actions in any networked environment.
To learn more, visit www.anutanetworks.com/ASA.
Additional Contributors: Manisha Dhan
