Today we continue our in-depth article series on server management. Gleb, the head of the DevOps department, tells how server monitoring is carried out and why it is so important.
Why server monitoring is important
If you do not monitor your server, you will never know whether your store operates properly or not. Unfortunately, store owners get around to thinking about server issues too late, usually when users inform them about that. It definitely spoils the company image and reputation, besides the store incurs losses when it’s down.
This being said, when launching a commercial project it’s crucial to take care of server monitoring — a special system that will keep you in the loop of things. You’ll know exactly what happened, what caused the issue, and how to fix it to prevent it in the future.
It’s especially frustrating to deal with issues that could have been avoided. Say, you’re running an ad campaign, but your server fails you—it can’t handle the load. But you knew that you could have increased server capacities in advance. (Don’t make me say I told you so.) What’s more, the disk space is running low or the SSL certificate expires at the wrong time. Pretty disappointing, huh?
If you have no idea what’s going on on the server, it becomes a more difficult task for developers to investigate server failures. We can only guess whether the load increased before the issue occurred, whether there was free disk space, whether the disk was overloaded with writing or reading, or whether the network was blinking.
Monitoring will help developers see a retrospective of server performance. Diagnosing problems becomes easier, and issues are resolved in shorter time frames. Monitoring lets you promptly predict the traffic growth or reduction, as well as spot bottlenecks.
How we carry it out
We offer the monitoring service to clients of the technical support service. Our team monitors server health across all the important metrics. We work on 12-hour shift schedules to ensure your website is up 24/7.
See how the desktop with the monitored metrics looks like.
As soon as a failure is detected, our engineer will see it on a big screen. At this very moment he will get a voice call alert informing him about malfunctions. The bot will keep calling until the engineer picks up and listens through the whole message.
Unless the issue is resolved within 30 minutes, another administrator will be notified and get involved right away. The message will go up to the department management. Thus, we can be sure that no incident is missed either day or night.
We prefer not to disturb you with loads of notifications, but you can choose to receive instant email or SMS alerts anyway. Once we resolve the issue, we will analyze the reasons why it occurred and provide reports. To ensure that we get a good night’s sleep, we apply only sure-fire solutions to fight issues.
Therefore, we identify database issues, CPU overload, excessive memory usage, when the web server is down and other troubles in time. We know exactly when it’s time to extend resources. We work in a team and help developers effectively investigate and eliminate failures using server metrics.
We will help you take your project to the next level. Drop us a line at firstname.lastname@example.org regarding monitoring and notifications.