This repository has been archived on 2022-10-07. You can view files and clone it, but cannot push or open issues or pull requests.
skynet-webportal/setup-scripts/log-checker.py

123 lines
4.3 KiB
Python
Executable File

#!/usr/bin/env python3
import discord, sys, traceback, io, os
from bot_utils import setup, send_msg, sc_precision
from datetime import datetime, timedelta
from subprocess import Popen, PIPE
"""
log-checker checks the docker logs for siad.
Arguments:
1. path to a .env file (default is none so env variables can already be
preset)
2. docker container name name (default: "sia")
3. number of hours to look back in log (default: 1 hour)
"""
# The default check interval in hours.
DEFAULT_CHECK_INTERVAL = 1
bot_token = setup()
client = discord.Client()
@client.event
async def on_ready():
await run_checks()
await client.close()
async def run_checks():
print("Running Skynet portal log checks")
try:
await check_docker_logs()
except: # catch all exceptions
trace = traceback.format_exc()
await send_msg(client, "```\n{}\n```".format(trace), force_notify=False)
# check_journal checks the journal
async def check_journal():
print("\nChecking journal...")
# Get the systemd service name as an argument, or use "siad" as default.
service_name = "siad"
if len(sys.argv) > 2:
service_name = sys.argv[2]
# Get the systemd service name as an argument, or use "siad" as default.
check_interval = timedelta(hours=DEFAULT_CHECK_INTERVAL)
if len(sys.argv) > 3:
check_interval = timedelta(hours=int(sys.argv[3]))
now = datetime.now()
time = now - check_interval
time_string = "{}-{}-{} {}:{}:{}".format(time.year, time.month, time.day, time.hour, time.minute, time.second)
# Open the journal.
proc = Popen(["journalctl", "--user-unit", service_name, "--since", time_string], stdin=PIPE, stdout=PIPE, stderr=PIPE, text=True)
std_out, std_err = proc.communicate()
if len(std_err) > 0:
await send_msg(client, "Error reading journalctl output: {}".format(std_err), force_notify=True)
return
# If there are any critical errors. upload the whole log file.
if "Critical" in std_out or "panic" in std_out:
upload_name = "{}-{}-{}-{}-{}:{}:{}.log".format(service_name, time.year, time.month, time.day, time.hour, time.minute, time.second)
await send_msg(client, "Critical error found in log!", file=discord.File(io.BytesIO(std_out.encode()), filename=upload_name), force_notify=True)
return
# No critical errors, return a heartbeat type message
pretty_before = time.strftime("%I:%M%p")
pretty_now = now.strftime("%I:%M%p")
await send_msg(client, "No critical warnings in log from `{}` to `{}`".format(pretty_before, pretty_now))
# check_docker_logs checks the docker logs by filtering on the docker image name
async def check_docker_logs():
print("\nChecking docker logs...")
# Get the container name as an argument or use "sia" as default.
container_name = "sia"
if len(sys.argv) > 2:
container_name = sys.argv[2]
# Get the container id for siad.
stream = os.popen('docker ps -q --filter name=^{}$'.format(container_name))
image_id = stream.read().strip()
# Get the number of hours to look back in the logs or use 1 as default.
check_hours = DEFAULT_CHECK_INTERVAL
if len(sys.argv) > 3:
check_hours = int(sys.argv[3])
now = datetime.now()
time = now - timedelta(hours=check_hours)
time_string = "{}h".format(check_hours)
# Read the logs.
proc = Popen(["docker", "logs", "--since", time_string, image_id], stdin=PIPE, stdout=PIPE, stderr=PIPE, text=True)
std_out, std_err = proc.communicate()
if len(std_err) > 0:
await send_msg(client, "Error reading docker logs output: {}".format(std_err), force_notify=True)
return
# If there are any critical errors. upload the whole log file.
if "Critical" in std_out or "panic" in std_out:
upload_name = "{}-{}-{}-{}-{}:{}:{}.log".format(container_name, time.year, time.month, time.day, time.hour, time.minute, time.second)
await send_msg(client, "Critical error found in log!", file=discord.File(io.BytesIO(std_out.encode()), filename=upload_name), force_notify=True)
return
# No critical errors, return a heartbeat type message
pretty_before = time.strftime("%I:%M%p")
pretty_now = now.strftime("%I:%M%p")
await send_msg(client, "No critical warnings in log from `{}` to `{}`".format(pretty_before, pretty_now))
client.run(bot_token)