---
title: at — one-time scheduled execution
date: 2026-10-18
time: 12:00
model: admin
category: programs
author_type: human
summary: A command for scheduling a job to run only once at a specific time. Useful for maintenance and deployment windows.
tags: program, Automation and scheduled tasks, at
---

**💬 "at command"**

The at command is a scheduled job that runs **only once**. Unlike cron, it does not repeat.

Use it to schedule restarts or maintenance during low-traffic times.

- **Concept**: Runs a command once at a specified time.
- **Time specification**: Use the format '02:30' or 'now + 10 minutes'.
- **List**: View scheduled jobs with atq.
- **Cancel**: Cancel a scheduled job with atrm.
## Commonly used options

| Option | Description |
| --- | --- |
| at <time> | Accepts a job to run at the specified time in input mode |
| atq | View scheduled job list |
| atrm <number> | Cancel a scheduled job |
| -f <file> | Schedule a script file |

## 3 essential command combinations

- `echo '/opt/app/bin/restart.sh' | at 03:30` — Schedule a restart at 03:30 in the early morning
- `atq` — View scheduled job list
- `atrm 7` — Cancel scheduled job number 7
**💬 "Show me an example that also considers preventing the same problem from recurring"**

Common situations on production servers are organized in order.

### Step 1 · Schedule

Specify the time and command.

```
echo '/opt/app/bin/restart.sh > /tmp/restart.log 2>&1' | at 03:30
```

### Step 2 · Verify

View the scheduled job list.

```
atq
```

### Step 3 · Cancel/Validate

Cancel if necessary and check the result log.

```
atrm 7 ; cat /tmp/restart.log
```

## In practice, it is used like this

### 1. Restart during low-traffic time

**Situation** — You need to restart the service during a weekly maintenance window.

```
echo 'systemctl restart myapp' | at 04:00
```

**Result** — A one-time restart runs at 04:00.

**Point** — Include a health check after restart in the same script for automatic verification.

### 2. Schedule a temporary job

**Situation** — You want a one-time cleanup 30 minutes later.

```
echo '/opt/app/bin/cleanup.sh' | at now + 30 minutes
```

**Result** — It runs once 30 minutes later.

**Point** — If the atd service is stopped, the job will not run, so check it.

## Things to watch out for

- If the atd service is stopped, scheduled jobs will not run.
- Without output redirection, you cannot see the results.
- You need to verify whether jobs persist across reboots (in some environments, atd alone is not enough).
## Actual execution screen

This is the result after scheduling a restart job at 04:00 in the early morning and checking the list.

![at command execution result: one-time scheduled job and list](https://games.aidebatehub.com/assets/at-demo.png)

Production server

---
Source (HTML): https://games.aidebatehub.com/programs/linux/automation/at.html
This file is an auto-generated Markdown twin of the DevHub document.