R12.1-JA-2025June05

Part III: Problem Diagnosis and Preventive Automation

Network problems are often organized by a Ticket System in the form of incidents. In the real world, 95% of network problems are repetitive. However, these identical or similar problems are diagnosed the same way each time without automation. Further, over 50% of problems are preventable, caused by misconfiguration, performance degradation or security violation. But enterprises lack the automation to enforce the design rules, best practice, or security policy to prevent these problems from happening again effectively.

NetworkBrain Problem Diagnosis Automation System (PDAs) is the intent-based hybrid network automation platform to address both problems by automating the diagnosis of the repetitive problem and the enforcement of preventive measures across the entire network. It can address up to 95% of network service tickets, reduce service times by up to half, and proactively prevent up to 50% of abnormal network conditions that would otherwise lead to the creation of even more service tickets.

null

From the end user's perspective, the output of PDAs is NetworkBrain Incident Pane/Portal, a central collaboration platform for troubleshooting and data sharing for each problem.
 Diagram Description automatically generated

The underlying system has three essential flows, as shown in the following system architecture diagram:

  • Automation Creation Flow: where diagnosis know-how is turned into automation assets across the entire network in the form of Network Intent (NI) or Executable Runbook (RB) inside the no-code platform.
  • Automation Installation Flow: where various automation assets are connected to future problem diagnosis through Trigger from the ticket system, human interaction, or NetworkBrain’s adaptive monitoring system.  
  • Automation Execution Flow: where automation is executed in response to an external symptom in three successive methods, namely triggered, interactive, and preventive. All execution output is organized inside the NetworkBrain incident pane for each distinctive Incident.

Network Intent (NI) is the core of PDAs. NI describes a network design for a specific network device, what these design baselines are like, and how to verify the design works properly. It provides a way to document network design intent, allowing other engineers to understand the device's design and baseline or normal state of a particular device. NI is supported by the Visual Parser on Configurations, SNMP, CLI commands, API in no-code automation flow.

NI is the cornerstone of the Problem Diagnosis Automation System (PDAS) reference workflow. The following diagram illustrates how intent works in PDAS system:

PDAS-NG has updated no-code intent capture, replication, and execution in many ways, including:
  • On-Demand replication of Intent
  • Major enhancement of Intent’s no-code programmability.
  • Use Intent to drive programable notification to 3rd party solution.
  • Other enhancements such as supporting intent across API-based network such as SDN or Cloud
Enables intent to address five of most complex use cases for network management:
  • Outage Prevention
  • Transient Problems Troubleshooting
  • Application Performance
  • Network Security
  • Protective Change

This chapter will elaborate on PDAs from the following aspects: 



Kubernetes Visual Manager Automation Basis