Dynamic Telemetry Cadence Control for Cloud Network Fabrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional telemetry data collection in networks places a significant load on network elements and central controllers, leading to inefficiencies and delays due to fixed frequency collection, which does not adapt to network needs, and results in unnecessary resource use and difficulty in managing and retrieving data subsets.
Innovation Solution
Implementing a cloud-based network controller that uses doubly-indexed state blocks with cursors to manage telemetry data, allowing dynamic adjustment of collection frequency based on network conditions and witness requirements, and synchronizing data efficiently without buffering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed frequency telemetry data collection is used, then network elements and controllers have simple data collection mechanism, but data processing load increases and resource usage becomes inefficient
Solution Approach 1:
The patent implements dynamic data collection frequency adjustment based on network conditions. The system transitions from fixed frequency collection to variable frequency collection where the collection cadence is adjusted according to network state, enabling efficient resource utilization while maintaining simple collection mechanisms.
Solution Approach 2:
The system changes the time parameter of data collection from fixed intervals to variable intervals based on network conditions. By modifying the collection frequency parameter dynamically, the system achieves efficient resource usage without complicating the fundamental data collection approach.
2Device complexity
If fixed frequency data collection is implemented, then data collection is straightforward, but network changes are not captured timely
Solution Approach 1:
The system dynamically adjusts data collection frequency based on detected network conditions. When network changes are detected or anticipated, the collection frequency increases to capture changes timely, while maintaining simple collection mechanisms during stable conditions.
Solution Approach 2:
The system incorporates feedback mechanisms where network state information is used to adjust collection frequency. This feedback loop enables timely detection and capture of network changes while keeping the overall collection approach simple and straightforward.
3Speed
If high frequency data collection is used to capture rapid changes, then network changes are captured timely, but processing load on controllers increases significantly
Solution Approach 1:
The system dynamically adjusts collection frequency based on actual network conditions. During stable periods, collection occurs at low frequency to minimize processing load, while switching to high frequency only when network changes are detected or anticipated, thus balancing speed and processing requirements.
Solution Approach 2:
The system changes the time parameter of data collection from fixed high frequency to variable frequency based on network state. This parameter adjustment enables timely change capture when needed while significantly reducing processing load during stable conditions.
4Quantity of substance
If frequent data collection is performed, then data availability is high, but unnecessary resource consumption increases
Solution Approach 1:
The system implements dynamic data collection frequency adjustment where high data availability is maintained during critical periods through increased collection frequency, while resource consumption is reduced during stable periods through lowered frequency, thus balancing availability and efficiency.
Solution Approach 2:
The system changes the collection frequency parameter dynamically to match actual data availability needs. This parameter adjustment ensures sufficient data availability when needed while minimizing unnecessary resource consumption during stable network conditions.
Data Source
AI summary
Described herein are devices, systems, methods, and processes for managing the collection and synchronization of telemetry data in a network overseen by a cloud-based network controller. This can be achieved by representing telemetry data as doubly-indexed state blocks within a shared meta-schema. Each type within the schema may be associated with a temporal list of objects of that type, providing ordered indexing by name and by time of last change. Cursors representing data witnesses may be threaded in place within these lists, enabling synchronization of telemetry data between devices without buffering. The system can dynamically adjust telemetry collection cadence in real time across devices in the fabric as users navigate the user interface. This approach can provide an effective mechanism to manage the load created by the telemetry, particularly in the context of network switches and telemetry collection.