Skip to content

fix(cdi): bound refresh restart attempts - #2118

Open
kvnloo wants to merge 3 commits into
NVIDIA:mainfrom
kvnloo:fix/cdi-refresh-bounded-retries
Open

kvnloo wants to merge 3 commits into
NVIDIA:mainfrom
kvnloo:fix/cdi-refresh-bounded-retries

Conversation

@kvnloo

@kvnloo kvnloo commented Oct 1, 2026

Copy link
Copy Markdown

Summary

The packaged CDI refresh service already gets TimeoutStartSec=90s from 10-container-engines.conf. This keeps that timeout policy and makes the restart-rate window long enough that the sixth start remains inside the same window even when every failed attempt consumes the full timeout.

Changes

  • keep StartLimitBurst=5
  • use StartLimitInterval=10min
  • preserve packaged TimeoutStartSec=90s
  • use the backward-compatible [Service] spelling
  • assert effective timeout, burst, and interval in E2E coverage
  • run the assertion hermetically in normal Go CI

Validation

Green downstream:

  • build
  • lint/check
  • unit tests
  • CodeQL
  • hermetic CDI refresh restart-limit regression

systemd v219 and current source were checked for directive compatibility. All commits are DCO-signed.

Fixes #2110

Signed-off-by: Kevin Rajan <veeman961@gmail.com>
Signed-off-by: Kevin Rajan <7121943+kvnloo@users.noreply.github.com>
Signed-off-by: Kevin Rajan <7121943+kvnloo@users.noreply.github.com>
@copy-pr-bot

copy-pr-bot Bot commented Oct 1, 2026

Copy link
Copy Markdown

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

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

Labels

None yet

Projects

None yet

1 participant