OCI audit logs export

README.md

Tryout https://docs.oracle.com/en/learn/oci-audit-logs/#introduction in 2025-09-09

# Text search and replace COMPARTMENT_ID with your compartment-id

# Your oci-cli must be fully configured to run this python script!
# e.g. test getting a sample audit log on a specific day (page 0):
oci logging-search search-logs --search-query 'search "COMPARTMENT_ID/_Audit" | sort by datetime desc' --time-start 2025-08-01"T00:00:00Z" --time-end 2025-08-01"T23:59:00Z"

# Install https://github.com/astral-sh/uv
curl -LsSf https://astral.sh/uv/install.sh | sh

# relogin to your term then setup a new python project
uv init oci-audit-export
cd oci-audit-export

# properly add the python dependencies as they were available in 2022-12
uv add pandas==1.5.3 numpy==1.24.3

# copy over the main.py file from the link above or the updated one attached (prints which commands are executed to stdout)

# then execute the python script like:
# all 3 params are mandadory (even if you configured a default compartment-id!)
uv run main.py -s 2025-08-01 -e 2025-08-01 -c COMPARTMENT_ID

main.py

# (c) Maninder Singh Flora (Cloud Architect) https://docs.oracle.com/en/learn/oci-audit-logs/#introduction
# Mario Ranftl 2025-09-09 Adapted to print subshell command to stdout before execution

import pandas as pd
import csv
import os
import datetime
from datetime import timedelta,date
import subprocess
import json
import sys
import getopt

def myfunc(argv):

    START_DATE = ""
    END_DATE = ""
    COMPARTMENT_ID = ""
    arg_help = "\nUsage:\n\npython3 main.py -s <log_start_date in YYYY-MM-DD Format> -e <log_end_date in YYYY-MM-DD Format> -c <Compartment_ID>\n".format(argv[0])
    try:
        opts, args = getopt.getopt(argv[1:],"hs:e:c:")
    except:
        print("\nenter valid argument values. Try -h option with script to know the usage.\n")
        sys.exit(2)

    for opt, arg in opts:
        if opt in ("-h"):
            print(arg_help)  # print the help message
            sys.exit(2)
        elif opt in ("-s"):
            START_DATE = arg
        elif opt in ("-e"):
            END_DATE = arg
        elif opt in ("-c"):
            COMPARTMENT_ID = arg

    INC = -1
    year1, month1, day1 = map(int, START_DATE.split('-'))
    DATE1 = datetime.date(year1, month1, day1)
    year2, month2, day2 = map(int, END_DATE.split('-'))
    DATE2 = datetime.date(year2, month2, day2)
    NUM_DAYS = DATE2 - DATE1
    NUM_DAYS = NUM_DAYS.days

# Function for converting JSON to .CSV format

    def convert_csv():

        with open('out.json', 'r') as file:
            data = json.load(file)
        datetimez = []
        compartmentID = []
        compartmentName = []
        message = []
        tenantId = []
        userAgent = []
        path = []
        ingestedtime = []
        type = []
        id = []
        for ele in data['data']['results']:
            datetimez.append(ele['data']['datetime'])
            compartmentID.append(ele['data']['logContent']['data']['compartmentId'])
            compartmentName.append(ele['data']['logContent']['data']['compartmentName'])
            message.append(ele['data']['logContent']['data']['message'])
            tenantId.append(ele['data']['logContent']['data']['identity']['tenantId'])
            userAgent.append(ele['data']['logContent']['data']['identity']['userAgent'])
            path.append(ele['data']['logContent']['data']['request']['path'])
            ingestedtime.append(ele['data']['logContent']['oracle']['ingestedtime'])
            type.append(ele['data']['logContent']['type'])
            id.append(ele['data']['logContent']['id'])
        finaldate = []
        for ts in datetimez:
            finaldate.append(datetime.datetime.fromtimestamp(int(ts) / 1000).strftime('%Y-%m-%d %H:%M:%S'))

        output = zip(finaldate, compartmentID, compartmentName, message, tenantId, userAgent, path, ingestedtime, type, id)
        output = list(output)
        df = pd.DataFrame(output)
        df.to_csv('/tmp/out.csv', header=False , mode='a',index=False)
        return None

# Check and validate the .CSV file in the /tmp directory

    os.system("touch /tmp/out.csv" )
    os.remove("/tmp/out.csv")
    header=['Date-Time', 'CompartmentID', 'CompartmentName', 'Message', 'TenantId', 'UserAgent', 'Path', 'Ingested-Time', 'Type', 'ID']
    data = []
    with open('/tmp/out.csv', 'a') as f:
        writer = csv.writer(f)
        writer.writerow(header)
        writer.writerows(data)

# Block for saving Audit Logs in JSON format to out.json file    

    for i in range(INC, NUM_DAYS):
        print("\ncollecting logs for", DATE1)
        cmd1 = ''' oci logging-search search-logs --search-query 'search "''' + str(COMPARTMENT_ID) + '''/_Audit" | sort by datetime desc' --time-start ''' + str(DATE1) + '''"T00:00:00Z" --time-end ''' + str(DATE1) + '''"T23:59:00Z" > out.json '''
        print(f"Executing: {cmd1}")
        p = subprocess.Popen(cmd1, shell=True, stdout=subprocess.PIPE)
        (output, err) = p.communicate()
        convert_csv()
        cmd2 = ''' oci logging-search search-logs --search-query 'search "''' + str(COMPARTMENT_ID) + '''/_Audit" | sort by datetime desc' --time-start ''' + str(DATE1) + '''"T00:00:00Z" --time-end ''' + str(DATE1) + '''"T23:59:00Z" | grep "opc-next-page" | awk -F":" '{print $2}' | tr -d '"' | tr -d " " | tail -1 '''
        print(f"Executing: {cmd2}")
        PAG = subprocess.check_output(cmd2, shell=True).strip().decode('ascii')

        while (PAG != ""):
            cmd3 = ''' oci logging-search search-logs --search-query 'search "''' + str(COMPARTMENT_ID) + '''/_Audit" | sort by datetime desc' --time-start ''' + str(DATE1) + '''"T00:00:00Z" --time-end ''' + str(DATE1) + '''"T23:59:00Z" --page ''' + str(PAG) + '''  > out.json '''
            print(f"Executing: {cmd3}")
            p = subprocess.Popen(cmd3, shell=True, stdout=subprocess.PIPE)
            (output, err) = p.communicate()
            convert_csv()
            cmd4 = ''' oci logging-search search-logs --search-query 'search "''' + str(COMPARTMENT_ID) + '''/_Audit" | sort by datetime desc' --time-start ''' + str(DATE1) + '''"T00:00:00Z" --time-end ''' + str(DATE1) + '''"T23:59:00Z" --page ''' + str(PAG) + ''' | grep "opc-next-page" | awk -F":" '{print $2}' | tr -d '"' | tr -d " " | tail -1 '''
            print(f"Executing: {cmd4}")
            PAG = subprocess.check_output(cmd4, shell=True).strip().decode('ascii')
        print("successfully collected logs for", DATE1)
        DATE1 += timedelta(days=1)
        i = i + 1

    print("\nThe .csv file is saved in location /tmp/out.csv")

if __name__ == "__main__":
    myfunc(sys.argv)