Facebook
Twitter
websavers
Regular Pleskian
Plesk: Version 17.8.11 Update #39, last updated on Feb 6, 2019 04:18 PM
OS: CentOS 6.10 (Final)
We've customized our server health monitor usage limits across the board. They generally work great. In all other cases when I've adjusted the thresholds, it has worked perfectly.
On at least one server we're regularly getting alerts that the alarm level has changed like this:
Server health parameter "Services > Apache memory usage" changed its status from "green" to "yellow".
Occasionally it reaches red.
Our Apache memory usage alerts are configured as follows:
Yellow: 60%
Red: 70%
The server health monitor in Plesk shows that the limit is 8GB memory and it shows the current amount and percentage accurately -- usually around 20% (currently 17.8%). When I click to show the graph for Apache Memory Usage, it shows that the highest peak is 2.4GB for today. 2.4/8GB = 30% memory usage.
When I hover over the (?) beside apache memory usage, it verifies that the thresholds are 60 and 70%.
Yet I've received multiple alerts today about it reaching the yellow and occasionally the red threshold.
I've pressed the button to detect hardware changes multiple times in case it has failed to do so (though it *does* accurately show the total memory as 8GB in the GUI).
Any ideas what's happening here? My best guesses are one of the following:
1. The notification system is not using the correct 8GB total memory to calculate its percentages
2. The notification system is not using the correct memory usage number to calculate its percentages
Where can I look to investigate further?
OS: CentOS 6.10 (Final)
We've customized our server health monitor usage limits across the board. They generally work great. In all other cases when I've adjusted the thresholds, it has worked perfectly.
On at least one server we're regularly getting alerts that the alarm level has changed like this:
Server health parameter "Services > Apache memory usage" changed its status from "green" to "yellow".
Occasionally it reaches red.
Our Apache memory usage alerts are configured as follows:
Yellow: 60%
Red: 70%
The server health monitor in Plesk shows that the limit is 8GB memory and it shows the current amount and percentage accurately -- usually around 20% (currently 17.8%). When I click to show the graph for Apache Memory Usage, it shows that the highest peak is 2.4GB for today. 2.4/8GB = 30% memory usage.
When I hover over the (?) beside apache memory usage, it verifies that the thresholds are 60 and 70%.
Yet I've received multiple alerts today about it reaching the yellow and occasionally the red threshold.
I've pressed the button to detect hardware changes multiple times in case it has failed to do so (though it *does* accurately show the total memory as 8GB in the GUI).
Any ideas what's happening here? My best guesses are one of the following:
1. The notification system is not using the correct 8GB total memory to calculate its percentages
2. The notification system is not using the correct memory usage number to calculate its percentages
Where can I look to investigate further?
