When running long-term bash scripts that output to log files, one critical piece of missing information is often the execution timestamp. Consider this common logging approach:

#!/bin/bash
while true; do
    echo "Processing data batch..."
    some_command >> /var/log/script_output.log
    sleep 60
done

Without timestamps, troubleshooting becomes challenging when you need to correlate events with system logs or identify performance patterns.

Here are three robust methods to add timestamps, each with different advantages:

1. Using the date Command Directly

The simplest method uses command substitution:

#!/bin/bash
echo "$(date) - Starting script execution" >> /var/log/script_output.log

For more precise formatting (ISO 8601 standard):

echo "$(date +"%Y-%m-%dT%H:%M:%S%z") - Event occurred" >> /var/log/script_output.log

2. Creating a Logging Function

For reusable timestamped logging throughout your script:

#!/bin/bash

log() {
    local timestamp=$(date +"%Y-%m-%d %H:%M:%S")
    echo "[${timestamp}] $1" >> /var/log/script_output.log
}

log "Initializing components"
# Your script logic here
log "Process completed successfully"

3. Using exec and Process Substitution

For comprehensive timestamping of all output:

#!/bin/bash

exec > >(while read -r line; do
    echo "$(date +"%F %T") $line" >> /var/log/script_output.log
done) 2>&1

echo "This line will be timestamped"
some_command
another_command

For production systems, consider these enhancements:

#!/bin/bash

LOG_FILE="/var/log/script_output.log"
MAX_LOG_SIZE=10485760 # 10MB

rotate_logs() {
    if [ -f "$LOG_FILE" ]; then
        local size=$(stat -c%s "$LOG_FILE")
        if [ "$size" -gt "$MAX_LOG_SIZE" ]; then
            mv "$LOG_FILE" "${LOG_FILE}.$(date +%Y%m%d%H%M%S)"
            touch "$LOG_FILE"
        fi
    fi
}

log() {
    rotate_logs
    local timestamp=$(date +"%Y-%m-%d %H:%M:%S.%3N")
    printf "[%s] %s\n" "$timestamp" "$1" >> "$LOG_FILE"
}

log "Script started with PID $$"

This implementation adds log rotation and millisecond precision timestamps.

For system-wide logging integration:

#!/bin/bash

log_system() {
    logger -t "myscript[$$]" "$1"
}

log_system "This goes to syslog with automatic timestamps"

---

When running long-term bash scripts that output to log files, missing timestamps makes it nearly impossible to:

• Track when events occurred
• Calculate time between operations
• Correlate with other system logs

Your current approach:

script.sh >> /var/log/logfile

Produces ambiguous logs like:

The server has booted up. Hmmph.
Connection timeout occurred
Service restarted

Modify your script to include timestamps for each output line:

#!/bin/bash
echo "$(date '+%a %b %d %T %Z %Y') The server has booted up. Hmmph."
# Other commands...

This produces properly formatted output:

Sat Sep 10 21:33:06 UTC 2011 The server has booted up. Hmmph.

For cleaner code, create a logging function:

#!/bin/bash

log() {
    echo "$(date '+%Y-%m-%d %H:%M:%S') $1" >> /var/log/logfile
}

log "Starting backup process"
# Your script commands...
log "Backup completed successfully"

For systems with moreutils installed:

script.sh | ts '%Y-%m-%d %H:%M:%S' >> /var/log/logfile

For high-frequency scripts:

echo "$(date '+%Y-%m-%d %H:%M:%S.%N') Event occurred"

Example output:

2023-11-15 14:22:31.548392100 Event occurred

When implementing timestamps, consider adding log rotation:

#!/bin/bash
LOG_FILE="/var/log/logfile_$(date +%Y-%m-%d).log"
log() {
    echo "$(date '+%Y-%m-%d %H:%M:%S') $1" >> "$LOG_FILE"
}