- Creating profiles
- Editing profiles
- Custom monitors
- Shell scripts
- Modifying scripts
- Silent servers
- Using clusters
- Viewing logs
To avoid confusion about functionality, it is important to understand when Pulse sends out notifications related to violated server monitoring thresholds.
Please note that this section does not apply to silent servers. For details on how Pulse handles silent servers, read the relevant support page.
Pulse always logs all incoming data, and updates your server's status accordingly, making it easy for you to observe your server's state at a given time. However, it does not respond blindly to the latest data it receives.
To understand why this is the case, consider the following example:
You've made some changes to your app and need to update your dependencies. You deploy these revisions to your server, and it complies, however during this time, your server's CPU spikes to 100% because it is busy performing tasks.
If Pulse responded blindly, it would tell you there is a problem with your server, when in reality, there isn't. It's just busy, it's not really under strain, or at least, it is only briefly so. The above example is just one of countless possible scenarios your server might encounter, and in order to be a good citizen, Pulse has to accommodate that reality.
A broader perspective
In order to avoid these false positives, Pulse uses an average of the last ten minutes. If your server CPU has been consistently reported as being above your threshold for those ten minutes, Pulse will notify you. Otherwise, it will assume the surge was a momentary issue and won't bother you about it.
This doesn't totally mitigate the initial problem... e.g. an intensive task could take longer than ten minutes, but it should help to address the vast majority of false positives, while also keeping you informed if your server truly is under strain.