Hardware Management/FMFM: Difference between revisions

From OpenCompute
Jump to navigation Jump to search
Line 12: Line 12:


===Scope===
===Scope===
* Coming Soon
The FMFM is a workstream about standardization of Fleetscale Memory Fault Management
 
*Proposed topics:
<ol>
<li>Standardize vendor agnostic architecture for memory error handling</li>
<ol style="list-style-type: lower-alpha; padding-bottom: 0;">
  <li style="margin-left:2em">Modularization of inputs from different hardware vendors</li>
  <li style="margin-left:2em; padding-bottom: 0;">APIs and connections between different modules from different vendors.</li>
  <li style="margin-left:2em; padding-bottom: 0;">Define the output of each module (failure cause, health information, RAS actions, etc.)</li>
</ol>
</li>
<li>Standardize memory error telemetry</li>
  <ol style="list-style-type: lower-alpha; padding-bottom: 0;">
  <li style="margin-left:2em">Format content for better fleet scale RAS management</li>
  <li style="margin-left:2em; padding-bottom: 0;">Troubleshooting, FRU replacement policies, etc.</li>
</ol>
</li>
<li>Coordinate with the broader OCP group to make sure there is a path for this general architecture</li>
</ol>


===Documents===
===Documents===

Revision as of 14:36, 18 July 2023

Welcome to the OCP Fleetscale Memory Fault Management (FMFM) WIKI

Fleetscale Memory Fault Management is a Worksteam within the Hardware Management Project.

Hardware Management Project


Leadership

Scope

The FMFM is a workstream about standardization of Fleetscale Memory Fault Management

  • Proposed topics:
  1. Standardize vendor agnostic architecture for memory error handling
    1. Modularization of inputs from different hardware vendors
    2. APIs and connections between different modules from different vendors.
    3. Define the output of each module (failure cause, health information, RAS actions, etc.)
  2. Standardize memory error telemetry
    1. Format content for better fleet scale RAS management
    2. Troubleshooting, FRU replacement policies, etc.
  3. Coordinate with the broader OCP group to make sure there is a path for this general architecture

Documents

  • Coming Soon


Past Fleetscale Memory Fault Management Events

  • Coming Soon