Skip to content

Report power state transitions to MGS#2065

Merged
hawkw merged 2 commits intomasterfrom
eliza/mgs-transitions
Jun 6, 2025
Merged

Report power state transitions to MGS#2065
hawkw merged 2 commits intomasterfrom
eliza/mgs-transitions

Conversation

@hawkw
Copy link
Member

@hawkw hawkw commented May 12, 2025

Basically, this commit glues together the changes from #2064 and
oxidecomputer/management-gateway-service#390. The
SpHandler::set_power_state implementation for compute sleds in the control-plane-agent
task now report the outcome of power state transitions using the new gateway_messages::PowerStateTransition type added in oxidecomputer/management-gateway-service#390.

This fixes oxidecomputer/management-gateway-service#270, and closes oxidecomputer/management-gateway-service#271.

We'll need to update MGS proper (in the Omicron repo) to pick up the new gateway-messages changes, as well.

Fixes #1466

Basically, this commit glues together the changes from
#2064 and
oxidecomputer/management-gateway-service#390. The
`SpHandler::set_power_state` implementation for compute sleds in the
`control-plane-agent`
task now report the outcome of power state transitions using the new
`gateway_messages::PowerStateTransition` type added in oxidecomputer/management-gateway-service#390.

This fixes oxidecomputer/management-gateway-service#270, and closes
oxidecomputer/management-gateway-service#271.

We'll need to update MGS proper (in the Omicron repo) to pick up the new
`gateway-messages` changes, as well.
@hawkw hawkw requested review from jgallagher and mkeeter May 12, 2025 21:00
Copy link
Collaborator

@mkeeter mkeeter left a comment

Choose a reason for hiding this comment

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

I didn't test this on hardware, but it looks reasonable to me!

@jgallagher jgallagher added this to the 16 milestone May 12, 2025
@jgallagher
Copy link
Contributor

Yep, LGTM too. Tagged this as R16 since we need to ship an MGS that knows how to receive this response in R15 first to keep wicket happy.

@hawkw
Copy link
Member Author

hawkw commented May 12, 2025

Yup, that's right. Unfortunately untangling the various knots necessary to pick up the MGS changes in Omicron will require untangling an entire stack of my nonsense, but what can you do?

@hawkw
Copy link
Member Author

hawkw commented May 12, 2025

(by which i mean: i'm gonna go do that. in case that wasn't clear)

hawkw added a commit to oxidecomputer/omicron that referenced this pull request May 14, 2025
This commit updates the MGS repo git dependencies to
oxidecomputer/management-gateway-service@0e1e055.
This picks up the following commits:

- oxidecomputer/management-gateway-service#389: Clap bump
- oxidecomputer/management-gateway-service#390: Make no-op
  `SetPowerState` calls succeed; needed to unblock
  oxidecomputer/hubris#2065 merging for R16
- oxidecomputer/management-gateway-service#388: Remove an unused flag
  bit from the ereport protocol
- oxidecomputer/management-gateway-service#391: Adds debug impls for the
  new `PowerStateTransition` type added in
  oxidecomputer/management-gateway-service#390

Most of these commits are mine, and should be relatively low risk to
merge --- trust me. 😇
@hawkw hawkw enabled auto-merge (rebase) June 6, 2025 21:05
@hawkw hawkw merged commit 893e85d into master Jun 6, 2025
135 checks passed
@hawkw hawkw deleted the eliza/mgs-transitions branch June 6, 2025 21:16
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.

SingleSp::set_power_state should probably be a bit more idempotent gimlet-seq-server: Allow A0 -> A0 and A2 -> A2 transitions

3 participants