When monitoring kernel logs through dmesg or /var/log/syslog, you've probably encountered lines like:

[ 1234.567890] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 1234.567899] last message repeated 3 times

While this compression saves disk space, it's problematic for:

• Real-time monitoring scripts
• Precise event timestamping
• Log analysis pipelines

This behavior originates from the kernel's printk subsystem, specifically controlled by:

/proc/sys/kernel/printk_devkmsg

The values represent:

• ratelimit: Default compression behavior
• ratelimit_burst: Number of repeats before compression kicks in

Add this to /etc/sysctl.conf:

kernel.printk_devkmsg = on

Then apply with:

sudo sysctl -p

Alternatively for immediate effect:

echo "on" | sudo tee /proc/sys/kernel/printk_devkmsg

For specific monitoring scripts, prepend:

echo 1 | sudo tee /proc/sys/kernel/printk_devkmsg
# Your monitoring command here
echo 0 | sudo tee /proc/sys/kernel/printk_devkmsg

After changes, test with:

for i in {1..5}; do logger "Test message $i"; done
tail -n 10 /var/log/syslog

Note that some logging daemons (like rsyslog) may still perform their own deduplication. Check their configuration if needed.

For cases where you can't modify system settings:

#!/bin/bash
dmesg --follow | while read -r line; do
  if [[ $line =~ "last message repeated" ]]; then
    count=$(echo $line | grep -oP '(?<=repeated )\d+')
    for ((i=0; i<count; i++)); do
      echo "${last_line}"
    done
  else
    last_line="$line"
    echo "$line"
  fi
done

---

When monitoring kernel logs on Debian/Ubuntu servers, you'll often encounter lines like:

kernel: [12345.67890] CPU1: Package temperature above threshold
kernel: [12345.67891] last message repeated 3 times

The rsyslog daemon (or syslog-ng in some cases) automatically compresses identical consecutive messages to save disk space. While useful for production systems, it's problematic for:

• Real-time monitoring scripts
• Debugging time-sensitive issues
• Log analysis tools expecting 1:1 message sequences

For systems using rsyslog (default on Debian/Ubuntu):

sudo nano /etc/rsyslog.conf

Add or modify these directives:

# Disable repeated message reduction
$RepeatedMsgReduction off

# For older rsyslog versions (<8)
$RepeatedMessagesReduction off

Then restart the service:

sudo systemctl restart rsyslog

Add this to your GRUB configuration (/etc/default/grub):

GRUB_CMDLINE_LINUX_DEFAULT="... log_buf_len=1M printk.always_kmsg_dump=y"

Update GRUB and reboot:

sudo update-grub
sudo reboot

For already compressed logs, use this awk script to expand repeats:

awk '{
  if (/last message repeated ([0-9]+) times/) {
    count = $4;
    while (count-- > 0) print prev_line;
  } else {
    print;
    prev_line = $0;
  }
}' /var/log/kern.log

Test your configuration by generating repeat messages:

for i in {1..5}; do
  logger -k "Test repeated message"
done
tail -n 10 /var/log/kern.log

You should now see five individual entries instead of one with a repeat count.

Disabling message compression will:

• Increase log file sizes by 10-30% in most cases
• Require more frequent log rotation
• Impact disk I/O during high-volume logging

For high-traffic servers, consider rate-limiting instead:

$SystemLogRateLimitInterval 0
$SystemLogRateLimitBurst 0