DevToolsmith

Cron Job Every Other Day

Standard cron can’t express “Every other day” in a single expression. Here is the closest schedule, 0 0 */2 * *, plus the workaround that fills the gap and copy-ready examples.

Runs 100% in your browser — nothing is uploaded

Cron schedule

Schedule time zone

UTC

Closest cron expression

0 0 */2 * *

At 00:00, every 2 days in a month. On its own, it runs on 186 days a year.

*/2 in the day-of-month field means odd days: the 1st, 3rd, 5th … 31st. So at the end of a 31-day month, the job runs on the 31st and again on the 1st, two days in a row. The count restarts every month.

Workaround: run a job every other day

For a strict 48-hour rhythm, run daily and skip every second day, using the day count since the Unix epoch:

crontab
0 0 * * * [ $(( $(date +\%s) / 86400 \% 2 )) -eq 0 ] && /usr/local/bin/job.sh

The same idea works on every scheduler below: keep the base schedule and put the check at the start of the job’s command or script.

Field by field

FieldValueMatches
minute 0 0
hour 0 0
day of month */2 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31
month * every value
day of week * every value

Examples

crontab

Run crontab -e and add:

0 0 */2 * * /usr/local/bin/job.sh >> /var/log/job.log 2>&1

For the exact schedule, use the workaround line above instead.

Kubernetes CronJob

apiVersion: batch/v1
kind: CronJob
metadata:
  name: my-job
spec:
  schedule: "0 0 */2 * *"
  timeZone: "Etc/UTC"          # Kubernetes 1.27+
  concurrencyPolicy: Forbid    # skip a run if the last one is still going
  jobTemplate:
    spec:
      template:
        spec:
          containers:
            - name: job
              image: ghcr.io/acme/job:1.0
          restartPolicy: OnFailure

GitHub Actions

on:
  schedule:
    - cron: '0 0 */2 * *'   # UTC

GitHub evaluates schedules in UTC, and runs can start several minutes late at busy times.

systemd timer

# /etc/systemd/system/job.timer
[Unit]
Description=Run job.service every other day

[Timer]
OnCalendar=*-*-01/2 00:00:00
Persistent=true

[Install]
WantedBy=timers.target

Enable it with systemctl enable --now job.timer, and check the next runs with systemd-analyze calendar "*-*-01/2 00:00:00". Persistent=true runs a missed job after downtime, which cron can’t do.

AWS EventBridge

cron(0 0 1/2 * ? *)

EventBridge cron has six fields (the last is the year), uses ? in either the day-of-month or the day-of-week field, and numbers weekdays from 1 (Sunday) to 7. Rules run in UTC; EventBridge Scheduler also accepts a time zone.

Similar schedules: daily

Have a different schedule? Paste any expression into the Cron Expression Generator, or browse all cron examples.

Sources

Spotted a mistake? Email hello@devtoolsmith.com and we’ll fix it.

Frequently asked questions

How do I run a cron job every other day?

No single standard cron expression does this exactly. */2 in the day-of-month field means odd days: the 1st, 3rd, 5th … 31st. So at the end of a 31-day month, the job runs on the 31st and again on the 1st, two days in a row. The count restarts every month. For a strict 48-hour rhythm, run daily and skip every second day, using the day count since the Unix epoch: 0 0 * * * [ $(( $(date +\%s) / 86400 \% 2 )) -eq 0 ] && /usr/local/bin/job.sh