Rotate and Compress Application Logs on a Schedule

Schedule

Setup

1. 1

   Verify or install logrotate

   On Debian/Ubuntu run: apt-get install -y logrotate. On RHEL/CentOS: yum install -y logrotate. Most Linux distributions ship it by default. Confirm with: logrotate --version.

2. 2

   Create or review your logrotate config

   Drop a config file at /etc/logrotate.d/myapp (see example in the script section). Key directives: daily or weekly, rotate N (how many old copies to keep), compress and delaycompress (gzip all but the most recent rotated file), missingok (skip silently if log is absent), notifempty (skip empty logs), postrotate/endscript (send a signal to your app to reopen its log file handle).

3. 3

   Create the wrapper shell script

   Save the script below to /usr/local/bin/rotate-app-logs.sh and make it executable: chmod +x /usr/local/bin/rotate-app-logs.sh. The script runs logrotate, captures its exit code, pings the CronJobPro heartbeat on success, or reports failure on error.

4. 4

   Add the cron entry

   Run: crontab -e (as root or the appropriate user) and add the line: 0 2 * * * /usr/local/bin/rotate-app-logs.sh >> /var/log/rotate-app-logs-cron.log 2>&1. This runs every day at 2 AM and appends stdout and stderr to a local audit log.

5. 5

   Create a CronJobPro heartbeat monitor

   In your CronJobPro dashboard create a new Heartbeat monitor. Set the period to 24 hours and the grace period to 30 minutes (giving the job until 2:30 AM to ping in). Copy the generated URL (https://cronjobpro.com/ping/<token>) and paste it into your script where indicated. Configure alert channels such as email or Slack so you are notified immediately if the ping does not arrive.

The script

bash

#!/usr/bin/env bash
# /usr/local/bin/rotate-app-logs.sh
# Runs logrotate for your application and pings a CronJobPro heartbeat on success.
# Edit LOGROTATE_CONF and HEARTBEAT_URL before use.

set -euo pipefail

LOGROTATE_CONF="/etc/logrotate.d/myapp"
HEARTBEAT_URL="https://cronjobpro.com/ping/YOUR_TOKEN_HERE"
STATE_FILE="/var/lib/logrotate/myapp.status"
LOG_TAG="rotate-app-logs"

log() {
  echo "[$(date -u '+%Y-%m-%dT%H:%M:%SZ')] $*"
}

log "Starting log rotation for config: ${LOGROTATE_CONF}"

# Run logrotate verbosely; --state keeps rotation state separate from system default
if logrotate --verbose --state "${STATE_FILE}" "${LOGROTATE_CONF}" 2>&1; then
  log "logrotate completed successfully"
  EXIT_CODE=0
else
  EXIT_CODE=$?
  log "logrotate exited with code ${EXIT_CODE}"
fi

# Report outcome to CronJobPro
if [ "${EXIT_CODE}" -eq 0 ]; then
  # Success ping: tells CronJobPro the job ran and finished cleanly
  curl --silent --max-time 10 --retry 2 "${HEARTBEAT_URL}" > /dev/null
  log "Heartbeat ping sent to CronJobPro"
else
  # Failure ping: immediate alert without waiting for the period to expire
  curl --silent --max-time 10 --retry 2 "${HEARTBEAT_URL}/fail" > /dev/null
  log "Failure ping sent to CronJobPro (exit code ${EXIT_CODE})"
  exit "${EXIT_CODE}"
fi

# ---------------------------------------------------------------
# Example /etc/logrotate.d/myapp (adjust paths and directives)
# ---------------------------------------------------------------
# /var/log/myapp/*.log {
#     daily
#     rotate 14
#     compress
#     delaycompress
#     missingok
#     notifempty
#     create 0640 www-data adm
#     sharedscripts
#     postrotate
#         # Tell your app to reopen its log file (example: nginx)
#         # kill -USR1 $(cat /run/myapp/myapp.pid) 2>/dev/null || true
#     endscript
# }
# ---------------------------------------------------------------

Frequently asked questions

More recipes