Skip to content

ASIC Thermal Monitoring HLD#557

Merged
liat-grozovik merged 3 commits intosonic-net:masterfrom
padmanarayana:asic_thermal_hld
Oct 22, 2020
Merged

ASIC Thermal Monitoring HLD#557
liat-grozovik merged 3 commits intosonic-net:masterfrom
padmanarayana:asic_thermal_hld

Conversation

@padmanarayana
Copy link
Copy Markdown
Contributor

This document describes the high level design of a poller for ASIC thermal sensors.

@jleveque
Copy link
Copy Markdown
Contributor

@keboliu, @Junchao-Mellanox: Please review.


```

The ThermalBase() implementation for ASIC sensors will return the latest values exported by the SwitchOrch to the stateDB's TEMPERATURE_INFO table.
Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

looks strange to me, the current idea is that platform API responsible to get the thermal sensor info from the lower layer and populate it to DB, and state DB info will feed CLI/SNMP or telemetry etc. but here you are suggesting platform API to read sensor info from DB?

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The objective is to present ASIC internal sensors to any advanced thermal control algorithm. The reason it appears dissimilar to the standard platform sensors is because the intent was not to let pmon (or any platform daemon) make SAI calls (and create a new dependency) since this is a unique case. Hence letting the poller be part of SwitchOrch.

Copy link
Copy Markdown
Collaborator

@keboliu keboliu left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

as comments left above.

@jleveque
Copy link
Copy Markdown
Contributor

@keboliu: Please review responses.

@padmanarayana padmanarayana requested a review from keboliu October 8, 2020 20:09
@liat-grozovik
Copy link
Copy Markdown
Collaborator

@Junchao-Mellanox and @keboliu appreciate your feedback

  • Is there any reason to have an additional thermal monitoring while there is an already valid solution for thermal control and new CLIs for doing so?
  • If there is something missing I believe it should be added on top of the existing functionality and not create new and parallel one.

Copy link
Copy Markdown
Contributor

@jleveque jleveque left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good to me. Please wait for other reviewers.

@liat-grozovik liat-grozovik merged commit f73b167 into sonic-net:master Oct 22, 2020
@sujinmkang
Copy link
Copy Markdown
Collaborator

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants