docs(boot-interface): refresh the boot-interface operator guide#3270
Open
chet wants to merge 1 commit into
Open
docs(boot-interface): refresh the boot-interface operator guide#3270chet wants to merge 1 commit into
chet wants to merge 1 commit into
Conversation
Contributor
Author
|
@coderabbitai full_review, thanks! |
Contributor
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
WalkthroughThis PR updates NICo documentation for boot-interface pairing and fallback, DPU-mode configuration, boot-order verification and recovery, four-store troubleshooting, operator CLI behavior, and host-ingestion fields. ChangesBoot Interface Documentation
Estimated code review effort: 2 (Simple) | ~10 minutes Possibly related PRs
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
Contributor
|
🐇✨ ✅ Action performedFull review finished. |
🔍 Container Scan SummaryNo Grype artifacts were found to aggregate. |
|
🌿 Preview your docs: https://nvidia-preview-pull-request-3270.docs.buildwithfern.com/infra-controller |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/provisioning/boot-interfaces-and-dpu-modes.md`:
- Around line 39-40: The fallback description in the
boot-interfaces-and-dpu-modes docs is too broad and implies every Redfish path
retries with interface id, including the direct-to-BMC escape hatch. Narrow the
wording in the paragraph that mentions NICo, MAC, and interface id so it only
applies to targeted Redfish operations that go through the store lookup flow,
and keep the separate redfish escape-hatch paragraph clearly excluded from that
retry logic.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Enterprise
Run ID: 15298eb3-360b-492b-8f3b-7b997e20eaf6
📒 Files selected for processing (2)
docs/provisioning/boot-interfaces-and-dpu-modes.mddocs/provisioning/ingesting-hosts.md
Contributor
Author
|
@coderabbitai full_review, thanks! |
Contributor
|
✅ Action performedFull review finished. |
Contributor
There was a problem hiding this comment.
Actionable comments posted: 3
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/provisioning/boot-interfaces-and-dpu-modes.md`:
- Around line 289-305: Update the Mermaid diagram around the machine
boot-interface lineage to place expected_machines.host_nics outside the four
diagnostic stores and label it clearly as the declaration/input source. Keep
only predicted_machine_interfaces, machine_interfaces, retained_boot_interfaces,
and explored_endpoints in the four-store representation, and show the
declaration source feeding prediction creation without implying it appears in
the CLI view or divergence calculation.
- Line 281: Clarify the HTTP-boot re-assertion limit consistently in the
documentation: update the description near SetBootOrder and the corresponding
text in Section 8 to state explicitly whether there are three total re-assert
attempts or three retries, using identical wording in both locations.
- Line 210: Correct the grammar in the “direct-to-BMC escape hatch” paragraph by
changing “a BMC NICo doesn't manage yet” to “a BMC that NICo doesn't manage
yet,” without altering the surrounding guidance.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Enterprise
Run ID: 8ee9e71a-fd2c-4849-a412-decdf5f5e03d
📒 Files selected for processing (2)
docs/provisioning/boot-interfaces-and-dpu-modes.mddocs/provisioning/ingesting-hosts.md
✅ Files skipped from review due to trivial changes (1)
- docs/provisioning/ingesting-hosts.md
Contributor
Author
|
@coderabbitai full_review, thanks! |
Contributor
|
✅ Action performedFull review finished. |
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/provisioning/boot-interfaces-and-dpu-modes.md`:
- Around line 353-361: Update the “First stop — the four-store view” description
to define the divergent flag using the boot-selection signals: effective owned
selection, explored default, and primary prediction disagreeing on more than one
MAC. Replace the inaccurate wording that implies any store or secondary
interface MAC can trigger divergence, while retaining the note that retained
records do not count.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Enterprise
Run ID: 9c55b587-0178-40c1-8c0d-deeefa46c404
📒 Files selected for processing (2)
docs/provisioning/boot-interfaces-and-dpu-modes.mddocs/provisioning/ingesting-hosts.md
chet
force-pushed
the
gh-issue-3269
branch
2 times, most recently
from
July 14, 2026 17:25
fa90830 to
380d9e4
Compare
chet
changed the title
The boot-interface operator guide merged a couple of weeks ago and the tooling kept moving right behind it: the four-store troubleshooting CLI landed two days later, and `SetBootOrder` grew its check-first skip and the `WaitForHttpBootDeviceApplied` re-assert loop. This refresh brings the guide back in line with what's shipped and makes the data model visible at a glance instead of prose-only. - Section 1 now states the runtime payoff of `MachineBootInterface` being a pair: when NICo targets a boot operation at the BMC it tries the MAC first and, when it knows the full pair, retries with the interface id on any error -- which is what keeps boot configuration working when a NIC drops out of the BMC's inventory. - `machine boot-interfaces` joins the CLI reference and leads the troubleshooting section -- the four stores side by side, the effective selection, and the `divergent` flag, with guidance on when divergence is normal. - The "applying the boot order" walkthrough matches the current `SetBootOrder` flow: check-first (nothing is re-applied unless something drifted), the reverted-HTTP-boot-device re-assert via `machine_setup`, and the bounded retry budget that ends in manual intervention -- plus a troubleshooting row for that operator-visible error. - Section 7 reframes the data model as the four stores (`predicted_machine_interfaces` -> `machine_interfaces` -> `retained_boot_interfaces`, with the `explored_endpoints` default alongside) and adds a mermaid lineage diagram. - The `redfish` command group is called out as the direct-to-BMC bypass -- no stores consulted, no interface-id fallback -- so nobody reaches for it on a managed host by accident. - `ingesting-hosts.md` now documents `host_nics` and cross-links the guide (it never had, in either direction). - Tightened the prose throughout -- cut filler words (`that`, `actually`) and vague demonstratives in favor of the specific nouns they pointed at. - Aligned the guide to the NVIDIA documentation style guide -- title-case headings, no em dashes, no Latinisms (`e.g.`/`i.e.`/`via`), `see` -> `refer to`, temporal `once` -> `after`, and no contractions. This supports NVIDIA#3269 Signed-off-by: Chet Nichols III <chetn@nvidia.com>
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/provisioning/boot-interfaces-and-dpu-modes.md`:
- Line 215: Remove the blank line between the two blockquoted paragraphs in the
documentation so they remain contiguous and satisfy markdownlint MD028;
alternatively, move the second paragraph outside the blockquote as a normal
paragraph.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Enterprise
Run ID: ee03b24a-fc6c-4a0d-bc49-5fb226865c5c
📒 Files selected for processing (2)
docs/provisioning/boot-interfaces-and-dpu-modes.mddocs/provisioning/ingesting-hosts.md
🚧 Files skipped from review as they are similar to previous changes (1)
- docs/provisioning/ingesting-hosts.md
| | `boot-override clear <interface-id>` | `ClearMachineBootOverride` | Revert to the default PXE/cloud-init. | | ||
|
|
||
| > Setting the DPU-first boot order directly (by MAC) is also exposed as a one-off action through the web UI ([Section 5](#5-web-ui)). Under normal operation the machine-controller sets the boot order automatically during ingestion ([Section 6](#6-behind-the-scenes-how-a-boot-device-is-chosen-and-set)). | ||
|
|
Contributor
There was a problem hiding this comment.
📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win
Remove the blank line inside the blockquote.
markdownlint reports MD028 here. Keep the two quoted paragraphs contiguous, or make the second paragraph a normal paragraph outside the blockquote.
🧰 Tools
🪛 markdownlint-cli2 (0.23.0)
[warning] 215-215: Blank line inside blockquote
(MD028, no-blanks-blockquote)
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@docs/provisioning/boot-interfaces-and-dpu-modes.md` at line 215, Remove the
blank line between the two blockquoted paragraphs in the documentation so they
remain contiguous and satisfy markdownlint MD028; alternatively, move the second
paragraph outside the blockquote as a normal paragraph.
Sources: Path instructions, Linters/SAST tools
This was referenced Jul 17, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The boot-interface operator guide merged a couple of weeks ago and the tooling kept moving right behind it: the four-store troubleshooting CLI landed two days later, and
SetBootOrdergrew its check-first skip and theWaitForHttpBootDeviceAppliedre-assert loop. This refresh brings the guide back in line with what's shipped and makes the data model visible at a glance instead of prose-only.MachineBootInterfacebeing a pair: every targeted Redfish operation tries the MAC first and retries with the interface id on any error, which is what keeps boot configuration working when a NIC drops out of the BMC's inventory.machine boot-interfacesjoins the CLI reference and leads the troubleshooting section -- the four stores side by side, the effective selection, and thedivergentflag, with guidance on when divergence is normal.SetBootOrderflow: check-first (nothing is re-applied unless something drifted), the reverted-HTTP-boot-device re-assert viamachine_setup, and the bounded retry budget that ends in manual intervention -- plus a troubleshooting row for that operator-visible error.predicted_machine_interfaces->machine_interfaces->retained_boot_interfaces, with theexplored_endpointsdefault alongside) and adds a mermaid lineage diagram.redfishcommand group is called out as the direct-to-BMC escape hatch -- no stores consulted, no interface-id fallback -- so nobody reaches for it on a managed host by accident.ingesting-hosts.mdnow documentshost_nicsand cross-links the guide (it never had, in either direction).This supports #3269
This also supports #3644