Skip to content
ZedCloud

Cron jobs

A Cron job runs a command in a fresh container on a schedule. Add Nodes ▾ → Compute → Cron job and set:

Field Default Notes
Schedule */5 * * * * Five fields: minute, hour, day, month, weekday. @hourly, @daily, @weekly, @monthly also work.
Image alpine Any image with the tools you need, for example postgres:17-alpine for pg_dump
Command echo hello Run with sh -c

The panel reads the schedule back in plain English with the next three run times, so you can check it.

Link the cron job to a database, bucket or app and it gets the same variables an app would. For example, a nightly Postgres dump to object storage, with the job linked to both nodes:

Terminal window
apk add --no-cache aws-cli && \
pg_dump "$DATABASE_URL" | gzip | \
aws --endpoint-url "$S3_ENDPOINT" s3 cp - "s3://$S3_BUCKET/backups/db-$(date +%F).sql.gz"

Schedule: 0 3 * * *. Image: postgres:17-alpine.

nightly-backup

ok

last run 03:00

0 3 * * *

db

running

Postgres

DATABASE_URL

backups

running

Object storage

S3_*

A nightly backup job linked to the database it dumps and the bucket it writes to.

Limits

Section titled “Limits”
  • Each run gets 512 MB and 1 CPU and is stopped after 5 minutes.
  • The Overview tab shows the last run, its result and the last 4,000 characters of output.
  • Sleeping services the job is linked to are woken for the run.
  • Cron jobs don’t count towards your memory and CPU quota.

To read another app’s files (for example to back up its SQLite database), see Sharing an app’s data.