Cron Job Not Running? 7 Checks

Q: How do I check if a cron job ran?

Look in the cron log. On systemd distros, journalctl -u cron (Debian, Ubuntu) or journalctl -u crond (Fedora, RHEL) shows one CMD line for every command cron started, with the user and the time; on older systems grep CRON /var/log/syslog or /var/log/cron. A CMD line proves cron started the job. It says nothing about whether the job succeeded, which is why every job should also write its own log.

Q: Why does my cron job work manually but not in cron?

cron runs jobs with a minimal environment: PATH=/usr/bin:/bin, no ~/.bashrc, no exported variables, /bin/sh instead of bash unless SHELL= is set, and your home directory as the working directory. Commands installed in /usr/local/bin, ~/.local/bin or /snap/bin fail with command not found, and relative paths point somewhere else. Use absolute paths, set PATH= at the top of the crontab, and test with env -i PATH=/usr/bin:/bin HOME=$HOME /bin/sh -c 'your command'.

Q: Why does % break my cron job?

In a crontab line, an unescaped % ends the command and turns the rest into standard input, so tar czf backup-$(date +%F).tgz runs as tar czf backup-$(date + with the remainder fed to stdin. Escape every % as \% inside the crontab, or move the command into a script file, where % is an ordinary character.

Q: Do I need to restart cron after editing crontab?

No. crontab -e installs the new crontab and cron checks for changes every minute, so the edit takes effect within a minute. A restart is only needed if you edited files under /etc/cron.d with a tool that did not update the file's modification time. If a new job does not run, check that the minute you scheduled has actually passed and look for its CMD line in the log.

Build the line instead of debugging it

The Cron Job Builder writes the crontab line with logging, PATH, MAILTO and flock already in it, warns about the day-of-month/day-of-week trap, and exports the same schedule as a systemd timer. The Cron Wrapper Generator adds timeout, retry and alerts. Background on why cron fails quietly: Bash Scripts That Survive Cron.

A cron job that does not run produces no error message anywhere you would look. A job that runs and fails usually produces one, and cron throws it away. Either way you find out days later, when the backup you needed is not there. The causes are few and they are always the same seven, so check them in order instead of guessing.

How Do I Know Whether Cron Ran the Job at All?

First, is there a cron daemon? Then, did it start the job? cron writes one CMD line to its log for every command it starts:

bash

systemctl status cron            # crond on Fedora/RHEL/Arch (cronie)
journalctl -u cron --since today | grep CMD
grep CRON /var/log/syslog        # systems without journald

No CMD line for your job means cron never started it: the daemon is down, the line does not parse, or the schedule is not what you think. A CMD line means cron did its part and the job itself failed, which is checks 4 to 7.

Why Does the Schedule Not Fire When I Expect?

Five fields, then the command: minute hour day-of-month month day-of-week. Two traps cause most surprises. A line with only four schedule fields is not an error cron reports to you; it simply never runs. And when both day-of-month and day-of-week are restricted, cron runs on days matching either: 0 1 1 * 0 means the 1st of every month and every Sunday, not "the 1st, if it is a Sunday" (man 5 crontab). The systemd timers guide proves that one on this machine, and the Cron Job Builder warns about it as you build.

What Else Breaks a Job That Cron Did Start?

The script itself. It must exist at the exact absolute path, be executable (chmod +x), and have Unix line endings: a CRLF shebang fails with bad interpreter. Fix "Permission denied" and Fix bad interpreter cover both.
PATH. cron's PATH is /usr/bin:/bin. Anything in /usr/local/bin, ~/.local/bin or /snap/bin is command not found, exit 127. Use full paths or a PATH= line at the top of the crontab; Fix "bash: command not found" reproduces it.
%. cron turns an unescaped % into a newline and feeds the rest of the line to stdin, so date +%F breaks. Write \%, or put the command in a script.
The environment. No ~/.bashrc, no exported variables, /bin/sh unless SHELL= is set, working directory $HOME. Bash Environment Variables explains what a child process inherits and what cron does not give it.
Overlap. A job slower than its schedule starts a second copy on top of the first. flock -n skips the overlapping run; timeout stops a hung one. flock and timeout have the patterns.

Where Did the Error Message Go?

cron mails a job's output to its owner. With no mail transfer agent installed, which is the default on most desktops and many cloud images, it cannot, so it discards the output and logs one line. This machine has no MTA, and between 2026-09-09 (the oldest journal entry) and 2026-10-03 cron logged this 117 times, every one a job's output, possibly an error, thrown away:

text

2026-09-14T05:00:01-05:00 laptop CRON[pid]: (CRON) info (No MTA installed, discarding output)
2026-09-14T05:01:01-05:00 laptop CRON[pid]: (CRON) info (No MTA installed, discarding output)

(Hostname and PID replaced.) The fix costs nothing: end every job with >> /var/log/jobname.log 2>&1, so stdout and stderr land in a file. For alerts on failure, Send Slack Alerts From Bash or Send an Email Alert.

The Script

Save as cron-doctor.sh. It checks the daemon and mail delivery on this machine, then every job line of a crontab (yours by default, or a file you pass) for the problems above.

bash

#!/bin/bash
# Script: cron-doctor.sh
# Purpose: A cron job that never runs, or runs and fails, leaves no error anywhere you would look — this checks the daemon and every line of a crontab for the seven causes that account for most of them.
# Usage: ./cron-doctor.sh [CRONTAB_FILE]   (default: your own crontab, via crontab -l)
set -euo pipefail

CHECK="✓"
CROSS="✗"
WARN="!"

# cron's PATH when the crontab does not set one (Debian/Ubuntu cron and cronie both default to this).
CRON_DEFAULT_PATH="/usr/bin:/bin"
PROBLEMS=0
problem() { echo "  $CROSS $*"; PROBLEMS=$((PROBLEMS + 1)); }

# 1. Is a cron daemon running at all?
DAEMON=""
for unit in cron crond cronie; do
  if systemctl is-active --quiet "$unit" 2>/dev/null; then DAEMON="$unit"; break; fi
done
if [[ -n "$DAEMON" ]]; then
  echo "$CHECK cron daemon active ($DAEMON.service)"
else
  echo "$CROSS no active cron daemon (cron, crond or cronie): nothing in any crontab will run"
  PROBLEMS=$((PROBLEMS + 1))
fi

# 2. Can cron mail the output anywhere? Without an MTA, output that is not redirected is thrown away.
HAS_MTA=0
if command -v sendmail >/dev/null; then
  HAS_MTA=1
  echo "$CHECK an MTA is installed (sendmail found): unredirected output is mailed"
else
  echo "$WARN no MTA (no sendmail): output a job does not redirect is discarded"
  if [[ -n "$DAEMON" ]] && LOST=$(journalctl -u "$DAEMON" --since "-7 days" -o cat 2>/dev/null | grep -c "No MTA installed"); then
    echo "  cron logged \"No MTA installed, discarding output\" $LOST time(s) in the last 7 days"
  fi
fi

# 3. Read the crontab.
if [[ $# -ge 1 ]]; then
  CRONTAB=$(cat -- "$1")
  SOURCE="$1"
else
  CRONTAB=$(crontab -l 2>/dev/null) || { echo "$CROSS you have no crontab (crontab -l failed)"; exit 1; }
  SOURCE="crontab -l"
fi
echo "checking $SOURCE"
if grep -q $'\r' <<< "$CRONTAB"; then
  problem "the crontab has CRLF line endings: cron reads the carriage return as part of each command"
fi

CRONTAB_PATH=""
n=0
while IFS= read -r line || [[ -n "$line" ]]; do
  n=$((n + 1))
  line="${line%$'\r'}"
  [[ "$line" =~ ^[[:space:]]*(#|$) ]] && continue
  # Environment lines (PATH=, MAILTO=, SHELL=) apply to the jobs below them.
  if [[ "$line" =~ ^[[:space:]]*([A-Za-z_][A-Za-z0-9_]*)[[:space:]]*= ]]; then
    [[ "${BASH_REMATCH[1]}" == "PATH" ]] && CRONTAB_PATH="${line#*=}"
    continue
  fi

  read -r -a f <<< "$line"
  if [[ "$line" =~ ^[[:space:]]*@[a-z]+[[:space:]]+(.*)$ ]]; then
    dom="*"; dow="*"
    cmd="${BASH_REMATCH[1]}"
  elif [[ "$line" =~ ^[[:space:]]*([^[:space:]]+[[:space:]]+){5}(.*)$ ]]; then
    dom="${f[2]}"; dow="${f[4]}"
    cmd="${BASH_REMATCH[2]}"
  else
    echo "line $n: $line"
    problem "fewer than 6 fields: cron cannot parse this line"
    continue
  fi
  echo "line $n: $cmd"

  # Both day fields restricted: cron runs when EITHER matches (man 5 crontab).
  if [[ "$dom" != "*" && "$dow" != "*" ]]; then
    problem "day-of-month ($dom) and day-of-week ($dow) are both set: cron runs on either, not only when both match"
  fi

  # An unescaped % ends the command; everything after it becomes stdin.
  if grep -qE '(^|[^\\])%' <<< "$cmd"; then
    problem "unescaped % in the command: cron cuts the command there; write \\% (e.g. date +\\%F)"
  fi

  first="${cmd%% *}"
  if [[ "$first" == /* ]]; then
    if [[ ! -e "$first" ]]; then
      problem "$first does not exist"
    elif [[ ! -x "$first" ]]; then
      problem "$first is not executable: chmod +x $first"
    elif head -n 1 -- "$first" | grep -q $'\r'; then
      problem "$first has a CRLF shebang: it fails with bad interpreter"
    fi
  elif ! type -t "$first" >/dev/null 2>&1 || [[ "$(type -t "$first")" == "file" ]]; then
    # Found by your shell's PATH is not the same as found by cron's.
    if ! PATH="${CRONTAB_PATH:-$CRON_DEFAULT_PATH}" command -v "$first" >/dev/null; then
      problem "'$first' is not on cron's PATH (${CRONTAB_PATH:-$CRON_DEFAULT_PATH}): use its full path ($(command -v "$first" || echo 'not found here either'))"
    fi
  fi

  if (( ! HAS_MTA )) && ! grep -qE '>' <<< "$cmd"; then
    problem "output is not redirected and there is no MTA: errors vanish; append >> /path/job.log 2>&1"
  fi
done <<< "$CRONTAB"

if (( PROBLEMS )); then
  echo "$CROSS $PROBLEMS problem(s) found"
  exit 1
fi
echo "$CHECK no problems found"

Prerequisites

bash 4 or later and systemd for the daemon check (on systems without it, that check reports no daemon; read the rest). journalctl access to the cron unit needs membership of adm or systemd-journal, or sudo. No root otherwise.

What Does the Script Print?

Against a demonstration crontab with one healthy job and one example of each problem. The scripts live in a scratch directory, shown as /opt/jobs:

text

$ cat demo.crontab
# m h dom mon dow command
MAILTO=""
0 2 * * * /opt/jobs/backup.sh >> /var/log/backup.log 2>&1
30 6 * * 1-5 /opt/jobs/report.sh >> /tmp/report.log 2>&1
*/15 * * * * /opt/jobs/sync.sh >> /tmp/sync.log 2>&1
0 3 * * * /opt/jobs/cleanup.sh >> /tmp/cleanup.log 2>&1
0 4 * * * yt-dlp -U >> /tmp/ytdlp.log 2>&1
0 5 * * * tar czf /backup/home-$(date +%F).tgz /home >> /tmp/tar.log 2>&1
0 1 1 * 0 /opt/jobs/backup.sh --monthly >> /var/log/backup.log 2>&1
0 7 * * /opt/jobs/backup.sh
0 8 * * * /opt/jobs/backup.sh --quick
$ ./cron-doctor.sh demo.crontab
✓ cron daemon active (cron.service)
! no MTA (no sendmail): output a job does not redirect is discarded
  cron logged "No MTA installed, discarding output" 58 time(s) in the last 7 days
checking demo.crontab
line 3: /opt/jobs/backup.sh >> /var/log/backup.log 2>&1
line 4: /opt/jobs/report.sh >> /tmp/report.log 2>&1
  ✗ /opt/jobs/report.sh is not executable: chmod +x /opt/jobs/report.sh
line 5: /opt/jobs/sync.sh >> /tmp/sync.log 2>&1
  ✗ /opt/jobs/sync.sh has a CRLF shebang: it fails with bad interpreter
line 6: /opt/jobs/cleanup.sh >> /tmp/cleanup.log 2>&1
  ✗ /opt/jobs/cleanup.sh does not exist
line 7: yt-dlp -U >> /tmp/ytdlp.log 2>&1
  ✗ 'yt-dlp' is not on cron's PATH (/usr/bin:/bin): use its full path (~/.local/bin/yt-dlp)
line 8: tar czf /backup/home-$(date +%F).tgz /home >> /tmp/tar.log 2>&1
  ✗ unescaped % in the command: cron cuts the command there; write \% (e.g. date +\%F)
line 9: /opt/jobs/backup.sh --monthly >> /var/log/backup.log 2>&1
  ✗ day-of-month (1) and day-of-week (0) are both set: cron runs on either, not only when both match
line 10: 0 7 * * /opt/jobs/backup.sh
  ✗ fewer than 6 fields: cron cannot parse this line
line 11: /opt/jobs/backup.sh --quick
  ✗ output is not redirected and there is no MTA: errors vanish; append >> /path/job.log 2>&1
✗ 8 problem(s) found
exit=1

The first two lines are this machine's real state: cron is running, there is no MTA, and 58 job outputs were discarded in the past week. Line 3 is the healthy job. Running it with no argument checks your own crontab the same way. It exits 1 when it finds anything, so it can run from cron itself and write its report to a log.

How Do I Stop It Happening Again?

Absolute paths, PATH= at the top of the crontab, \% for every percent sign, >> log 2>&1 on every line, and flock -n plus timeout around anything that can run long. Or move the job to a systemd timer, which logs to the journal, catches up runs missed while the machine was off, and never overlaps itself; the Cron Job Builder exports any crontab line as a timer. The Production Bash Toolkit ships a cron wrapper that does the flock, timeout and logging for you.

Frequently Asked Questions

How do I check if a cron job ran?

Look in the cron log. On systemd distros, journalctl -u cron (Debian, Ubuntu) or journalctl -u crond (Fedora, RHEL) shows one CMD line for every command cron started, with the user and the time; on older systems grep CRON /var/log/syslog or /var/log/cron. A CMD line proves cron started the job. It says nothing about whether the job succeeded, which is why every job should also write its own log.

Why does my cron job work manually but not in cron?

cron runs jobs with a minimal environment: PATH=/usr/bin:/bin, no ~/.bashrc, no exported variables, /bin/sh instead of bash unless SHELL= is set, and your home directory as the working directory. Commands installed in /usr/local/bin, ~/.local/bin or /snap/bin fail with command not found, and relative paths point somewhere else. Use absolute paths, set PATH= at the top of the crontab, and test with env -i PATH=/usr/bin:/bin HOME=$HOME /bin/sh -c 'your command'.

Why does % break my cron job?

In a crontab line, an unescaped % ends the command and turns the rest into standard input, so tar czf backup-$(date +%F).tgz runs as tar czf backup-$(date + with the remainder fed to stdin. Escape every % as \% inside the crontab, or move the command into a script file, where % is an ordinary character.

Where does cron output go?

cron mails anything a job prints to the crontab's owner, or to MAILTO if set. If no mail transfer agent such as postfix or exim is installed, the mail cannot be sent and cron throws the output away, logging (CRON) info (No MTA installed, discarding output). Redirect every job explicitly with >> /var/log/job.log 2>&1, so errors land in a file you can read, or route them to an alert script.

Do I need to restart cron after editing crontab?

No. crontab -e installs the new crontab and cron checks for changes every minute, so the edit takes effect within a minute. A restart is only needed if you edited files under /etc/cron.d with a tool that did not update the file's modification time. If a new job does not run, check that the minute you scheduled has actually passed and look for its CMD line in the log.

Part of the bash snippets collection

Prevent Overlapping Cron Runs With flock — the overlap fix
The timeout Command — the hang fix
Fix "bash: command not found" — the PATH cause, reproduced
Bash Environment Variables — what cron's environment lacks

Cron Job Not Running? 7 Checks That Find the Cause

How Do I Know Whether Cron Ran the Job at All?

Why Does the Schedule Not Fire When I Expect?

What Else Breaks a Job That Cron Did Start?

Where Did the Error Message Go?

The Script

Prerequisites

What Does the Script Print?

How Do I Stop It Happening Again?

Frequently Asked Questions

How do I check if a cron job ran?

Why does my cron job work manually but not in cron?

Why does % break my cron job?

Where does cron output go?

Do I need to restart cron after editing crontab?

The Production Bash Toolkit

Get the bashlib starter

Related Snippets

Frequently Asked Questions

Cron Job Not Running? 7 Checks That Find the Cause

How Do I Know Whether Cron Ran the Job at All?

Why Does the Schedule Not Fire When I Expect?

What Else Breaks a Job That Cron Did Start?

Where Did the Error Message Go?

The Script

Prerequisites

What Does the Script Print?

How Do I Stop It Happening Again?

Frequently Asked Questions

How do I check if a cron job ran?

Why does my cron job work manually but not in cron?

Why does % break my cron job?

Where does cron output go?

Do I need to restart cron after editing crontab?

Related Scripts

The Production Bash Toolkit

Get the bashlib starter

Related Snippets

Frequently Asked Questions