fix(scheduler): reload is unverifiable and unknown YAML sections vanish silently

[mt-alarcon]

Two independent bugs in scheduler.py, both the same shape: the scheduler knows something the operator needs and drops it silently.

1. SIGHUP reload is unverifiable

print(f"  {ts} [reload] SIGHUP received — clearing schedule and re-reading routines")
print(f"  {ts} [reload] {total} routines scheduled")

No flush=True. Since stdout is normally redirected to a log file, Python uses block buffering and these lines sit in the buffer indefinitely — in practice [reload] never appears in the log at all.

The consequence is worse than a missing line. Someone checking "did my config reload?" sees nothing and cannot distinguish a working reload from a broken one. On our install this produced a real false conclusion in both directions: one person could not prove the reload had worked, and a second person read a pre-reload routine run as proof it had failed. The reload had worked all along; the evidence only surfaced when the process shut down and flushed its buffer.

The routine-execution print already passes flush=True. Only these two did not.

2. Unknown YAML sections vanish without a word

The loader reads exactly daily, weekly, monthly. Any other top-level key — a typo, or a section someone assumed was supported — is dropped silently: the routines under it simply never run. No error, no warning, nothing in the log.

We found a section that had been sitting unread for weeks with routines under it. Nothing in the system could have told us; it was found by reading the loader source.

Now the loader warns and names the offending sections. It does not add new sections or change what is read — it only stops the silence.

Scope

Deliberately minimal: two bug fixes, no new features, no behaviour change beyond the two messages. Both are running on a live install.

Related but not included, since it would be a feature rather than a fix: the monthly loop is hardcoded to day 1 / hour 8 and ignores any per-routine scheduling key. Upstream never advertised such a key, so that is a separate discussion.

Also available as a follow-up if wanted: a boot-time check that every scheduled routine's script actually exists, shouting on startup instead of failing once per tick forever. It caught four separate incidents downstream where a routine had never run once. It is left out here because our implementation leans on downstream-only alerting infrastructure and would need adapting.

🤖 Generated with Claude Code

Summary by Sourcery

Clarify scheduler behavior when reloading and reading YAML configuration so operators receive explicit feedback instead of silent failures.

Bug Fixes:

• Ensure SIGHUP-triggered reload messages are flushed to logs so reload events and resulting job counts are reliably visible.
• Emit warnings for unknown top-level YAML sections so misconfigured or unsupported routine groups no longer disappear silently.

[sourcery-ai]

Reviewer's guide (collapsed on small PRs)

Reviewer's Guide

Adds explicit logging and validation around scheduler reloads and YAML configuration sections to make operator-facing behavior observable and debuggable, while keeping functional behavior unchanged except for new messages.

Sequence diagram for YAML loader handling unknown sections

sequenceDiagram
    participant Scheduler
    participant YAMLConfig

    Scheduler->>YAMLConfig: _load_routines_from_yaml(schedule, config_path, is_plugin)
    YAMLConfig-->>Scheduler: config
    Scheduler->>Scheduler: determine plugin_slug
    Scheduler->>Scheduler: compute _unknown = [k for k in config if k not in _known_sections]
    alt unknown_sections_present
        Scheduler->>Scheduler: print(WARN unread section(s) ..., flush=True)
    end
    Scheduler->>Scheduler: iterate config.get(daily), config.get(weekly), config.get(monthly)

Loading

Sequence diagram for scheduler SIGHUP reload logging

sequenceDiagram
    actor Operator
    participant OS
    participant Scheduler

    Operator->>OS: send SIGHUP to scheduler
    OS->>Scheduler: deliver SIGHUP
    Scheduler->>Scheduler: shutdown(sig, frame)
    Scheduler->>Scheduler: _reload_flag.is_set()
    alt reload_flag_set
        Scheduler->>Scheduler: print([reload] SIGHUP received ..., flush=True)
        Scheduler->>Scheduler: schedule.clear()
        Scheduler->>Scheduler: setup_schedule()
        Scheduler->>Scheduler: total = len(schedule.get_jobs())
        Scheduler->>Scheduler: print([reload] total routines scheduled, flush=True)
    end
    Scheduler->>Scheduler: schedule.run_pending()

Loading

File-Level Changes

Change	Details	Files
Warn when the YAML scheduler config contains unknown top-level sections so operators know routines under those sections will never run.	Introduce a fixed set of known sections (daily, weekly, monthly) in the YAML loader. Compute the list of unknown top-level keys in the loaded config. Emit a WARN log message naming the unread sections and explaining that routines under them will never run, with flush enabled.	scheduler.py
Make SIGHUP-driven scheduler reloads verifiable by ensuring the reload log messages are flushed immediately.	Update the log print for receiving SIGHUP to pass flush=True. Update the log print reporting the number of routines scheduled after reload to pass flush=True.	scheduler.py

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[sourcery-ai]

Hey - I've reviewed your changes and they look great!

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}