As more and more people are putting their production apps on the Always Free Oracle Cloud, it's probably a good idea to start monitoring your instance. (but remember this is a no-SLA environment)
Over the last few weeks, it's happened to me - my EU test database was stopped and one of my compute instances was accidentally dropped and restored. Oracle notified me and took action themselves:
If I had monitoring on, I would have seen this myself. The above case I couldn't solve on my own, but it could happen that some software you are running goes down, or something else makes your site/app unavailable due to your own code. If that happens you want to take action and verify if things are ok.
There are many monitoring tools out there, some are paid, some are free. I've used Pingdom and UptimeRobot. It takes only a minute to set it up. I'll show how to do it with UptimeRobot.
Sign-up at uptimerobot.com
Click the New Monitor button and specify the URL you want to monitor.
Add a type of connection where you want to be notified and you are done!
You can simulate what happens if you go into your compute instance and stop the webserver.
nginx -s stop
Depending on the monitoring interval you will get an email within 5 minutes when your URL can't be reached.
You can also log in and see in the dashboard how much uptime you had and when you were down:
nginx -s stop
Depending on the monitoring interval you will get an email within 5 minutes when your URL can't be reached.
You can restart your web server by doing:
systemctl restart nginx
A few minutes later UptimeRobot will notify you all is good again :)
And surely in the dashboard, we are UP again:
Happy monitoring!
What caused the database to stop?
ReplyDelete