Skip to content

codeSplitters limits: try the queue action in a browser and on the live Worker #284

New issue
New issue
Closed as not planned
Closed as not planned
codeSplitters limits: try the queue action in a browser and on the live Worker#284
Assignees
QuinnsCode

Description

@QuinnsCode
QuinnsCode
opened on Oct 10, 2026

Follow-up from #280. The queue action is covered by a test, run in-process, but nobody has tried it in a browser or on the deployed Worker yet.

To check after the deploy that applies `migrations/0021_limit_jobs.sql`:

  • On `#/admin/limits`, "Queue it" is offered for Dig up a repo only, and saving it sticks.
  • Set digs to 1 per hour with queue, sign in as a non-admin, and dig twice. The second dig shows the "you're 1st in line" toast and the over-the-cap log shows "queued".
  • Reset that caller from the admin page. Within about 5 seconds the toast clears and the page opens the new repo.
  • Queue one more, close the tab, reset the caller, and confirm the five-minute cron makes the repo.

Remember that admins are never limited, so test as a non-admin.

🤖 Generated with Claude Code

Activity

  1. QuinnsCode
    self-assigned this
    on Oct 10, 2026

  2. QuinnsCode commented on Oct 10, 2026

    @QuinnsCode
    QuinnsCode
    on Oct 10, 2026
    OwnerAuthor

    Partway through, 2026-10-10:

    • ✅ Migration 0023_limit_jobs.sql (renumbered from 0021) is applied on live D1.
    • ✅ On #/admin/limits, "Queue it" is offered only on the queueable rules: Dig up a repo, plus Preview deploys and Dependency checks, which were added since. Setting digs to 1 per hour with queue saved and stuck after a reload.
    • ⏸ The non-admin steps are blocked by codeSplitters live: email sign-in (and sometimes sign-up, admin limits) hangs on the Worker #333: email sign-in hangs on the live Worker, and sign-up mostly does too. A fresh non-admin account made one dig fine, but the session couldn't be kept up to finish.

    The live dig rule is still set to 1 per hour with queue. Put it back to 5 per hour with reject on the admin page if this isn't picked up soon.

  3. QuinnsCode commented on Oct 10, 2026

    @QuinnsCode
    QuinnsCode
    on Oct 10, 2026
    OwnerAuthor

    Update: steps 2 and 3 now pass on live, as a non-admin email account:

    • Second dig: "Over the limit for dig up a repo: you're 1st in line, about 34 minutes." The over-the-cap log shows queued.
    • Reset from the admin page: the toast cleared and the page opened #/r/queuetest-mv2aejdk/q2-mv2aejdk 5.8s later.
    • Side note: the dig right after a replay wasn't queued. The replayed job doesn't count in the new window, so the cap allows one more.

    Left: the cron replay after the tab closes. It's blocked again by #333 (live sign-in hangs), which is labelled fable now. The live dig rule is still 1 per hour with queue. Put it back to 5 per hour with reject.

  4. QuinnsCode commented on Oct 11, 2026

    @QuinnsCode
    QuinnsCode
    on Oct 11, 2026
    OwnerAuthor

    Tracked in #609, the one hands-on checklist. The steps here still apply; tick them there.

    🤖 Generated with Claude Code

  5. QuinnsCode
    closed this as not plannedon Oct 11, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

No labels
No labels

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions