Inside Sklotopolis - A Development Blog
Apr 9, 2020 11:19:00 GMT
Wreck, tomservo, and 4 more like this
Post by Dominikk on Apr 9, 2020 11:19:00 GMT
Inside Sklotopolis - A Development Blog
Sklotopolis has always been using modern technology to provide both a fault tolerant and stable systems for our players. But of course even our system is not perfect at all, there are outages and other errors that happen even on Sklotopolis. The Sklotopolis network is a private cloud environment hosted in Frankfurt (Germany) and has access to modern DDOS protection, fault tolerant network connections and a battery backed up power supply. So in theory there isn't anything which can go wrong, is it? Well of course there is, hardware outages, errors in the Wurm server application and human errors like scripts which fill up the disk space can't be prevented completly.
For these cases we are using an enterprise monitoring system giving us insight in all our servers, as checking them by hand would be an impossible task, there are far too many hosts and services inside the Sklotopolis network as we could check them every few minutes.
This is where our monitoring system kicks in!
With the help of Monitoring with Icinga we can view all details on all our servers 24/7 auto updated. There is no manual task needed, everything is completly automated. As you can see in the picture below not just the state is monitored but also the delay between the connections to make sure there are no lags or connection drops generated by our environment.
Next we not only can investigate the connection, but we can view everything that's going on on the server, RAM usage, CPU load, Free Disk Space, Users logged in, Security Updates available and much more. The picture below gives you an insight into what we monitor on every our hosts to ensure the best possible availability and performance for our players.
Monitoring is a very important part of Sklotopolis if things go wrong we are informed via E-Mail and Smartphone about something bad happening. We don't just want to wait until something went wrong and players are contacting us, our goal is to already have an answer when players are contacting us about a problem, because we have known it just seconds after it happened. And this can only be achieved by using an enterprise monitoring system for our servers which is what we use Icinga for. I wouldn't want to run a server environment without detailed, fast and clear monitoring.
Icinga is free and open source and can be used by anyone.
Sklotopolis has always been using modern technology to provide both a fault tolerant and stable systems for our players. But of course even our system is not perfect at all, there are outages and other errors that happen even on Sklotopolis. The Sklotopolis network is a private cloud environment hosted in Frankfurt (Germany) and has access to modern DDOS protection, fault tolerant network connections and a battery backed up power supply. So in theory there isn't anything which can go wrong, is it? Well of course there is, hardware outages, errors in the Wurm server application and human errors like scripts which fill up the disk space can't be prevented completly.
For these cases we are using an enterprise monitoring system giving us insight in all our servers, as checking them by hand would be an impossible task, there are far too many hosts and services inside the Sklotopolis network as we could check them every few minutes.
This is where our monitoring system kicks in!
Icinga and Open Source monitoring system used by small and big companies all over the world.
With the help of Monitoring with Icinga we can view all details on all our servers 24/7 auto updated. There is no manual task needed, everything is completly automated. As you can see in the picture below not just the state is monitored but also the delay between the connections to make sure there are no lags or connection drops generated by our environment.
Next we not only can investigate the connection, but we can view everything that's going on on the server, RAM usage, CPU load, Free Disk Space, Users logged in, Security Updates available and much more. The picture below gives you an insight into what we monitor on every our hosts to ensure the best possible availability and performance for our players.
But sometimes things just go wrong and then we need to know what went wrong, Icinga provides very powerful tools to track the history so we not only can see all the incidents that happened in the last hours, days or even years but we also can monitor the exact values which the different monitored services had in the last time. This makes it possible for us to exactly track down why there was a certain lag spike, why a server might have crashed or if there was an attack onto our system. All that is displayed in fancy charts which can be accessed easily, giving even unexperienced people a good overview of what could have happened.
Monitoring is a very important part of Sklotopolis if things go wrong we are informed via E-Mail and Smartphone about something bad happening. We don't just want to wait until something went wrong and players are contacting us, our goal is to already have an answer when players are contacting us about a problem, because we have known it just seconds after it happened. And this can only be achieved by using an enterprise monitoring system for our servers which is what we use Icinga for. I wouldn't want to run a server environment without detailed, fast and clear monitoring.
Icinga is free and open source and can be used by anyone.