Merge pull request #346 from NebulousLabs/gollum_docker_logs

Gollum will now check docker logs instead of the journal.

setup-scripts/README.md

@@ -104,8 +104,8 @@ At this point we have almost everything running, we just need to set up your wal
   > `docker exec caddy caddy reload --config /etc/caddy/Caddyfile`
 - Restarting nginx gracefully after making changes to nginx configs
   > `docker exec nginx openresty -s reload`
-- Checking siad service logs (follow last 50 lines)
-  > `journalctl -f -n 50 --user-unit siad`
+- Checking siad service logs (last hour)
+  > `docker logs --since 1h $(docker ps -q --filter "name=^sia$")`
 - Checking caddy logs (for example in case ssl certificate fails)
   > `docker logs caddy -f`
 - Checking nginx logs (nginx handles all communication to siad instances)

setup-scripts/funds-checker.py

@@ -5,15 +5,22 @@ health-checker runs simple health checks on a portal node using the siad API and
 dispatches messages to a Discord channel.
 """
 
-import discord, traceback
+import discord, traceback, asyncio
 from bot_utils import setup, send_msg, siad, sc_precision
 
 bot_token = setup()
 client = discord.Client()
 
+
+async def exit_after(delay):
+    await asyncio.sleep(delay)
+    exit(0)
+
+
 @client.event
 async def on_ready():
     await run_checks()
+    asyncio.create_task(exit_after(30))
     await client.close()
 
 

setup-scripts/log-checker.py

@@ -1,82 +1,91 @@
 #!/usr/bin/env python3
 
-import discord, sys, traceback, io
-from bot_utils import setup, send_msg, sc_precision
-
+import discord, sys, traceback, io, os, asyncio
+from bot_utils import setup, send_msg
 from datetime import datetime, timedelta
 from subprocess import Popen, PIPE
 
 """
-log-checker checks journal logs for siad.
+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. systemd service name (default: "siad")
+    2. docker container name name (default: "sia")
 
-    3. number of hours to look back in log (used as --since value in journalctl
-    command) (default: 1 hour)
+    3. number of hours to look back in log (default: 1 hour)
 
 """
 
-DEFAULT_CHECK_INTERVAL = timedelta(hours=1)
+# The default check interval in hours.
+DEFAULT_CHECK_INTERVAL = 1
 
 bot_token = setup()
 client = discord.Client()
 
+
+async def exit_after(delay):
+    await asyncio.sleep(delay)
+    exit(0)
+
+
 @client.event
 async def on_ready():
     await run_checks()
+    asyncio.create_task(exit_after(30))
     await client.close()
 
 
 async def run_checks():
     print("Running Skynet portal log checks")
     try:
-        await check_journal()
+        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...")
+# 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 systemd service name as an argument, or use "siad" as default.
-    service_name = "siad"
+    # Get the container name as an argument or use "sia" as default.
+    container_name = "sia"
     if len(sys.argv) > 2:
-        service_name = sys.argv[2]
+        container_name = sys.argv[2]
 
-    # Get the systemd service name as an argument, or use "siad" as default.
-    check_interval = DEFAULT_CHECK_INTERVAL
+    # 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_interval = timedelta(hours=int(sys.argv[3]))
+        check_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)
+    time = now - timedelta(hours=check_hours)
+    time_string = "{}h".format(check_hours)
 
-    # Open the journal.
-    proc = Popen(["journalctl", "--user-unit", service_name, "--since", time_string], stdin=PIPE, stdout=PIPE, stderr=PIPE, text=True)
+    # 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 journalctl output: {}".format(std_err), force_notify=True)
+        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(service_name, time.year, time.month, time.day, time.hour, time.minute, time.second)
+        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 messagej
+    # 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)

setup-scripts/setup-docker-services.sh

@@ -17,16 +17,19 @@ sudo curl -L "https://github.com/docker/compose/releases/download/1.25.5/docker-
 sudo chmod +x /usr/local/bin/docker-compose
 docker-compose --version # sanity check
 
-# Create dummy .env file for docker-compose usage with veriables
+# Create dummy .env file for docker-compose usage with variables
 # * DOMAIN_NAME - the domain name your server is using ie. example.com
 # * EMAIL_ADDRESS - this is the administrator contact email you need to supply for communication regarding SSL certification
 # * HSD_API_KEY - this is auto generated secure key for your handshake service integration
 # * CLOUDFLARE_AUTH_TOKEN` - (optional) if using cloudflare as dns loadbalancer (need to change it in Caddyfile too)
 # * AWS_ACCESS_KEY_ID - (optional) if using route53 as a dns loadbalancer
 # * AWS_SECRET_ACCESS_KEY - (optional) if using route53 as a dns loadbalancer
+# * API_PORT - (optional) the port on which siad is listening, defaults to 9980
+# * PORTAL_NAME - the name of the portal, required by the discord bot
+# * DISCORD_BOT_TOKEN - required by the discord bot
 if ! [ -f /home/user/skynet-webportal/.env ]; then
     HSD_API_KEY=$(openssl rand -base64 32) # generate safe random key for handshake
-    printf "DOMAIN_NAME=example.com\nEMAIL_ADDRESS=email@example.com\nSIA_WALLET_PASSWORD=\nHSD_API_KEY=${HSD_API_KEY}\nCLOUDFLARE_AUTH_TOKEN=\nAWS_ACCESS_KEY_ID=\nAWS_SECRET_ACCESS_KEY=\n" > /home/user/skynet-webportal/.env
+    printf "DOMAIN_NAME=example.com\nEMAIL_ADDRESS=email@example.com\nSIA_WALLET_PASSWORD=\nHSD_API_KEY=${HSD_API_KEY}\nCLOUDFLARE_AUTH_TOKEN=\nAWS_ACCESS_KEY_ID=\nAWS_SECRET_ACCESS_KEY=\nPORTAL_NAME=\nDISCORD_BOT_TOKEN=\n" > /home/user/skynet-webportal/.env
 fi
 
 # Start docker container with nginx and client

setup-scripts/setup-health-check-scripts.sh

@@ -8,8 +8,8 @@ sudo apt-get -y install python3-pip
 pip3 install discord.py
 pip3 install python-dotenv
 
-fundsCheck="0 0,8,16 * * * /home/user/skynet-webportal/setup-scripts/funds-checker.py /home/user/.sia/sia.env"
-logsCheck="0 0,8,16 * * * /home/user/skynet-webportal/setup-scripts/log-checker.py /home/user/.sia/sia.env siad 8"
+fundsCheck="0 0,8,16 * * * /home/user/skynet-webportal/setup-scripts/funds-checker.py /home/user/skynet-webportal/.env"
+logsCheck="0 0,8,16 * * * /home/user/skynet-webportal/setup-scripts/log-checker.py /home/user/skynet-webportal/.env sia 8"
 
 (crontab -u user -l; echo "$fundsCheck" ) | crontab -u user -
 (crontab -u user -l; echo "$logsCheck" ) | crontab -u user -