APScheduler Cron Transactional Email with Rovela (2026): CronTrigger + Idempotency-Key per Window

APScheduler owns the schedule — Rovela delivers the mail

APScheduler (Advanced Python Scheduler) runs cron, interval, and date triggers inside a Python process — commonly BackgroundScheduler or AsyncIOScheduler with a CronTrigger. Unlike Celery Beat, it does not require a separate broker-backed Beat service (unless you add a job store). That fits digests whose body depends on live ORM/API state at fire time rather than content frozen earlier with Rovela scheduled_at.

Distinct from Celery Beat Transactional Email — this post is APScheduler in-process (or with its own job store), not Celery beat_schedule.

Rovela queues, DKIM-signs, and delivers when your job method calls POST https://api.rovela.dev/emails (Send Email API, Quick Start, Authentication). APScheduler owns the cadence and “is due?” logic. Your job method owns aggregation, secrets for sending_access, and a stable Idempotency-Key per schedule window. Rovela only delivers.

This guide wires APScheduler → job method → Rovela that:

1. Declares scheduler.add_job(send_daily_digest, CronTrigger(hour=9, minute=0, timezone="UTC"), id="daily-digest-email", replace_existing=True)
2. Keeps ROVELA_API_KEY (sending_access), MAIL_FROM, and recipients in server-only config — never in browsers, SPA bundles, or a public Git tree
3. Aggregates a digest / report for the current schedule window inside the job method
4. Sends via Rovela with a window-scoped Idempotency-Key
5. Survives retries, restarts, and multi-worker fleets without duplicate mail

Rovela does not create APScheduler schedules, workers, or storage. APScheduler owns schedule → enqueue/fire → execute. Your job method owns query → map fields → POST /emails. Rovela owns delivery.

Distinct from:

• Quartz Scheduler Transactional Email — Java Quartz
• node-cron Transactional Email — node-cron
• NestJS Schedule Cron Email — NestJS @Cron
• Laravel Task Scheduler Email — Laravel Schedule::
• Spring @Scheduled Email — Spring @Scheduled
• Resque-scheduler Email — Resque-scheduler
• RQ Scheduler Email — RQ Scheduler
• delayed_job Recurring Email — delayed_job
• Solid Queue Recurring Email — Solid Queue
• Hangfire Recurring Jobs Email — Hangfire
• Sidekiq-Cron Transactional Email — Sidekiq-Cron
• Celery Beat Transactional Email — Celery Beat
• Agenda.js Scheduled Jobs Email — agenda
• BullMQ Repeatable Email — BullMQ
• Temporal Schedule Email — Temporal
• Trigger.dev Cron Email — Trigger.dev
• Inngest Cron Transactional Email — Inngest
• Kubernetes CronJob Email — K8s CronJob
• Vercel Cron Digest Email — Vercel cron
• Scheduled Email Sending — scheduled_at

Also distinct from RQ Scheduler (Redis Queue) — APScheduler is not RQ.

What APScheduler schedules vs what Rovela sends

Concern	Owner
Job id, CronTrigger, timezone, jobstore	APScheduler
Process / async loop must stay running	Your Python host
Coalesce / max_instances	APScheduler job defaults
Server secrets for ROVELA_API_KEY (sending_access)	Process / worker env / secret store
Query DB / APIs for the window	Your job method
Render HTML + text	Your job method
POST /emails, DKIM, queue, bounces	Rovela
Idempotency-Key = apscheduler-daily-digest:{yyyy-mm-dd}	Your job method

A 200 from Rovela means accepted and queued, not “already in the inbox” (Quick Start).

APScheduler constraints that shape the email job

• Process must stay up (or use a persistent job store + a runner).

• Multiple processes each with BackgroundScheduler can double-fire — run one scheduler process or use a shared job store with appropriate locking.

• Set explicit id + replace_existing=True and timezone on CronTrigger.

• Use max_instances=1 / coalesce for digest jobs that must not overlap.

• Distinct from Celery Beat’s broker heartbeats — do not mix both for the same mail cron without clear ownership.

• Keep sending_access server-only. Store ROVELA_API_KEY in environment / vault — never in client apps or public Git (Authentication).

• Prefer HTTPS POST /emails. Outbound SMTP from a worker is a poor fit (Email API vs SMTP).

When to use APScheduler + POST /emails (not scheduled_at)

Use APScheduler → live send when:

• The HTML must include fresh metrics at fire time
• You already run APScheduler and want schedules next to the same workers
• You need cron-like digests without adopting a separate SaaS scheduler
• One (or a few) transactional messages go to ops / founder / customer addresses looked up at run time

Use Rovela scheduled_at when the body is already correct and you only need “deliver later” without a running scheduler (Scheduled Email Sending). Prefer Celery Beat when you already run Celery workers at scale; prefer APScheduler for a single Python service without a broker.

Prerequisites

1. A verified sending domain in Rovela. Unverified from domains are refused with 400 (Quick Start).
2. An API key shaped like re_…. Prefer sending_access for the worker (Authentication).
3. Python app with apscheduler installed and a long-lived process starting the scheduler.
4. Config: ROVELA_API_KEY, MAIL_FROM, DIGEST_TO.
5. HTTP client (urllib, httpx, or requests) for Rovela — keep the API key server-side.

Register the schedule

pip install apscheduler

from apscheduler.schedulers.background import BackgroundScheduler
from apscheduler.triggers.cron import CronTrigger
from digest import send_daily_digest, send_ops_pulse

scheduler = BackgroundScheduler(timezone="UTC")
scheduler.add_job(
    send_daily_digest,
    CronTrigger(hour=9, minute=0, timezone="UTC"),
    id="daily-digest-email",
    replace_existing=True,
    max_instances=1,
    coalesce=True,
)
scheduler.add_job(
    send_ops_pulse,
    CronTrigger(minute="*/15", timezone="UTC"),
    id="ops-pulse-email",
    replace_existing=True,
    max_instances=1,
    coalesce=True,
)
scheduler.start()
# keep the process alive (Flask/FastAPI lifespan, or a dedicated worker loop)

Job method that calls Rovela

import os, json, urllib.request
from datetime import datetime, timezone

def send_daily_digest():
    day = datetime.now(timezone.utc).strftime("%Y-%m-%d")
    idempotency_key = f"apscheduler-daily-digest:{day}"
    stats = {"signups": 12, "paid": 3, "incidents": 1}  # replace with ORM queries
    to = [x.strip() for x in os.environ["DIGEST_TO"].split(",") if x.strip()]
    subject = f"Acme daily digest — {day}"
    text = f"Signups: {stats['signups']}\nPaid: {stats['paid']}\nIncidents: {stats['incidents']}"
    html = (
        f"<h1>Daily digest — {day}</h1><ul>"
        f"<li>Signups: <b>{stats['signups']}</b></li>"
        f"<li>Paid: <b>{stats['paid']}</b></li>"
        f"<li>Open incidents: <b>{stats['incidents']}</b></li></ul>"
    )
    # never log ROVELA_API_KEY
    print({"msg": "apscheduler digest", "day": day, "idempotency_key": idempotency_key})
    result = send_transactional(to=to, subject=subject, html=html, text=text,
                                idempotency_key=idempotency_key,
                                tags={"kind": "apscheduler-daily-digest", "day": day, "platform": "apscheduler"})
    print({"msg": "rovela accepted", "emailId": result.get("id"), "day": day})

def send_transactional(*, to, subject, html, text, idempotency_key, tags):
    api_key = os.environ["ROVELA_API_KEY"]
    from_addr = os.environ["MAIL_FROM"]
    payload = json.dumps({"from": from_addr, "to": to, "subject": subject, "html": html, "text": text, "tags": tags}).encode()
    req = urllib.request.Request(
        "https://api.rovela.dev/emails", data=payload, method="POST",
        headers={"Authorization": f"Bearer {api_key}", "Content-Type": "application/json", "Idempotency-Key": idempotency_key},
    )
    with urllib.request.urlopen(req, timeout=30) as resp:
        return json.loads(resp.read().decode())

Window-scoped Idempotency-Key

Rovela’s Idempotency-Key (up to 256 characters) makes a retry of the same send within 24 hours return the original response instead of mailing again (Send Email API). That matters because:

• Retries re-enter the job after 5xx / timeouts
• Crash after Rovela 200 but before the job is marked succeeded can redeliver
• Queue lag can shift wall-clock — key off the business schedule window, not millisecond now
• Multi-tenant jobs need tenant ids in the key

Recommended keys:

Key	Use when
apscheduler-daily-digest:{yyyy-mm-dd}	One ops digest per UTC day
apscheduler-ops-pulse:{yyyy-mm-dd}T{HH}-{MM}	/15 or hourly crontab (bucket to the window)
apscheduler-tokyo-ops:{yyyy-mm-dd}	Local-TZ morning digest
apscheduler-weekly-report:{yyyy}-W{ww}	One founder report per ISO week
apscheduler-tenant-digest:{tenantId}:{yyyy-mm-dd}	Per-tenant job args

If you mail many personalized digests in one job, prefer per-recipient keys (or POST /emails/batch for up to 100 non-template messages — batch rejects template / topic_id by design).

Attach secrets and verify the schedule

1. Set ROVELA_API_KEY (sending_access), MAIL_FROM, DIGEST_TO on the process that runs the scheduler/worker.
2. Confirm the schedule entry is registered with the expected cron and timezone.
3. Confirm workers are online.
4. Wait for the next tick, or trigger once in non-prod while keeping the window key stable.
5. Check logs for idempotencyKey + Rovela emailId. Optionally correlate later with Rovela webhooks.

Prefer sending_access on production senders — never full_access. Never log API keys. Rotate Rovela by create-then-revoke.

Failure modes and how to handle them

Symptom	Likely cause	What to do
Cron never fires	Scheduler/worker down	Keep the scheduling process always on
Double digests	Multiple schedulers / replicas	Single leader or distributed lock / persistent store
Wrong wall time	Assumed UTC / wrong timezone	Set timezone explicitly
Duplicate digests after retries	Unstable key / keyed by ms now	Use window-scoped keys; Idempotency-Key collapses duplicates within 24h
Rovela 401	Bad / missing ROVELA_API_KEY	Fix server env; confirm sending_access
Rovela 400 naming a domain	MAIL_FROM domain not verified	Verify the domain under Rovela Domains
Rovela 429	Rate limit	Raise/retry; keep the same Idempotency-Key
Secrets in Git	Key committed	Keep only in env / vault

Production checklist

• Verified Rovela sending domain; MAIL_FROM matches it
• ROVELA_API_KEY is sending_access, in worker env / vault only
• Scheduler + workers always running as documented for APScheduler
• Documented cron + timezone
• Rovela send uses a window-scoped Idempotency-Key
• Logs show successful emailId without printing secrets
• Alerts on consecutive job failures / scheduler down
• Optional: Rovela webhooks for email.bounced / email.complained on digest recipients

FAQ

Can the scheduler call Rovela without a job method? No usefully — the scheduler only wakes your code. Your method must aggregate live data and keep sending_access local. Rovela does not host your scheduler or workers.

Should I use Rovela scheduled_at from the job? Usually no. If the scheduler already fires at send time with a live body, call POST /emails immediately. Use scheduled_at when you accept the message now and want Rovela to deliver later without another wake (Scheduled Email Sending).

What key scope does sending_access cover? Enough to queue mail via POST /emails / batch. Reading messages, rescheduling, and cancel need full_access or a dashboard session (Send Email API, Authentication). Keep the worker on sending_access.

Does a 200 mean the founder already got the digest? No — it means Rovela accepted and queued the message. Use GET /emails/:id, events, or webhooks for delivery state (Quick Start).

Why both scheduler retries and Rovela Idempotency-Key? The scheduler retries the job when your code fails. Rovela’s key collapses duplicate sends if the HTTP call succeeds twice in the same window (or times out after Rovela accepted). You want both layers.

Ship the recurring job; let Rovela deliver

Register the APScheduler schedule with an explicit cron and timezone, put live aggregation and POST /emails in the job method, scope Idempotency-Key to the schedule window (and tenant), and keep sending_access server-only. APScheduler wakes on time; Rovela delivers the transactional mail.