Data Center BMC Telemetry Streaming with Dynamic Parameter Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing telemetry systems in large-scale data centers are inefficient and inflexible, leading to excessive monitoring and storage of unnecessary data, high network bandwidth consumption, and costly firmware updates, which can result in processing delays and customer dissatisfaction.
Innovation Solution
Implementing a smart telemetry monitoring and streaming (STMS) system with a baseboard management controller (BMC) that dynamically configures monitoring parameters and data streaming based on configuration files generated by a policy command service, reducing storage and network bandwidth usage by only transmitting relevant data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If static telemetry mechanisms are used to monitor all parameters, then comprehensive monitoring coverage is achieved, but storage space and network bandwidth are excessively consumed
Solution Approach 1:
The patent extracts and transmits only the necessary telemetry parameters from the data center nodes to the external system, rather than transmitting all parameters. The BMC determines which parameters are necessary based on node health status, and only those selected parameters are transmitted, significantly reducing storage and network bandwidth consumption while maintaining effective monitoring coverage.
2Reliability
If static telemetry mechanisms monitor all parameters, then complete data collection is achieved, but processing efficiency decreases due to excessive data volume
Solution Approach 1:
The system extracts and transmits only the necessary telemetry parameters based on node health status and monitoring requirements. By filtering out unnecessary parameters before transmission, the data volume is significantly reduced, which directly improves processing efficiency while maintaining the ability to collect complete information about node health issues.
3Adaptability or versatility
If firmware changes are made to add new monitoring parameters, then monitoring flexibility is improved, but system complexity and cost increase
Solution Approach 1:
The patent implements a dynamic parameter monitoring mechanism where the BMC can flexibly determine and adjust which telemetry parameters to monitor and transmit based on real-time node health status and external system requirements. This dynamic approach eliminates the need for firmware changes to add new monitoring capabilities, as the system can adapt its monitoring scope through software configuration rather than hardware/firmware modifications.
4Speed
If all telemetry parameters are transmitted continuously, then real-time monitoring capability is achieved, but network bandwidth consumption increases
Solution Approach 1:
The system implements periodic transmission of telemetry parameters based on node health status changes rather than continuous transmission. The BMC monitors parameters continuously but only transmits data when health status changes or at scheduled intervals, achieving real-time monitoring capability while significantly reducing network bandwidth consumption compared to continuous transmission of all parameters.
Data Source
AI summary
A system receives, from a policy command service executing in a control plane of a distributed computing system, a configuration file defining parameters to be monitored at the computing node, a timing factor indicating when the parameters are to be monitored, and a threshold for when the parameters are to be sent from the computing node. Based on the configuration file, while the computing node is operational in the distributed computing system, the baseboard management controller is configured to monitor the computing node in accordance with the configuration file. The baseboard management controller sends, to a storage at the distributed computing system, the parameters defined in the configuration file for analysis of maintenance and health of the computing node.


