Zabbix 5.4.6. doesn't trigger disk space alerts for some machines on dashboard

[cmartinez127]

I'm monitoring many machines, these machines (CentOS and Debian) have installed zabbix agent 1 and 2, but as the title says some of them won't appear on dashboard when disk space is critically low (90%).

I'm using "Linux filesystems by Zabbix agent active", a default Zabbix template and as you can see in the following screenshoot, the marked with a "X" is the one isn't working for some machines. However, the marked in green (Disk space is low 80%) works for any machine, even the problematic ones!

It doesn't matter the OS, I have this problem on CentOS and Debian machines and happens with the zabbix agent 1 and 2. That's all I've dismissed.

The strangest thing is that the trigger works for most of the machines, but in a few ones won't work...

Please help me, I need to fix this urgently because some machines ran out of space and I wasn't able to detect it on time.




Example of a problematic machine:

As you can see it's 98% of disk space but zabbix only shows 80% alert on dashboard.





Example of a working machine:

[liujun0253]

Your changed statement requires that the file system cannot be occupied higher than the $VFS.FS. PUSED. .MAX. CRIT macro, and the free space of this file system is less than 5GB, will the alarm be triggered?
If you only require an alarm when the file system occupies a proportion, you can directly remove the statement after and

[cmartinez127]

It's working, for now.

There is still a small issue, as you can see in the following screenshot it says " Space used: *UNKNOWN* of *UNKNOWN* ". This worked before removing the statement after "and" as you recommended.
Example of working alert: " Space used: 699.57 GB of 749.51 GB "

Screenshot: https://i.imgur.com/7V9TKr2.png

I think that now it doesn't collect data to make the Graphs.

Is there a way to solve this? Solving this is optional, I can live without it, since the alarm pops out on the dasboard that's all I need, for which I am very grateful, thanks.

[cmartinez127]

I dicovered in "Latest Data" that the host from the original post receives the data I want, but won't show it at the dashboard. Why? All hosts have the same items and triggers, no custom ones.

Please I still requiere hep.

[liujun0253]

Capacity alarms need to meet two conditions: exceeding the capacity threshold and achieving a certain rate of capacity growth. It's possible that your capacity usage is high, but the growth rate is not significant, which is why some alarms may not have been triggered.

[liujun0253]

In the diagram, your capacity growth rate is slow. You can go and check the trigger expressions in the monitoring template.

[cmartinez127]

Hi, thanks!

How would you delete the growth rate condition from this trigger expression (it's the default trigger)? I just need to be notified when the capacity usage is high (at 90% in this case).

Default trigger I want to modify:
last(/Linux filesystems by Zabbix agent active/vfs.fs.size[{#FSNAME},pused])>{$VFS.FS.PUSED.MAX.CRIT:"{#FSNAME}"} and ((last(/Linux filesystems by Zabbix agent active/vfs.fs.size[{#FSNAME},total])-last(/Linux filesystems by Zabbix agent active/vfs.fs.size[{#FSNAME},used]))<5G or timeleft(/Linux filesystems by Zabbix agent active/vfs.fs.size[{#FSNAME},pused],1h,100)<1d)

EDIT:
I did this but not sure if it's correct due to I don't have experience with Zabbix syntax.

last(/Linux filesystems by Zabbix agent active/vfs.fs.size[{#FSNAME},pused])>{$VFS.FS.PUSED.MAX.CRIT:"{#FSNAME}"} and ((last(/Linux filesystems by Zabbix agent active/vfs.fs.size[{#FSNAME},total])-last(/Linux filesystems by Zabbix agent active/vfs.fs.size[{#FSNAME},used]))<5G)

[liujun0253]

Your changed statement requires that the file system cannot be occupied higher than the $VFS.FS. PUSED. .MAX. CRIT macro, and the free space of this file system is less than 5GB, will the alarm be triggered?
If you only require an alarm when the file system occupies a proportion, you can directly remove the statement after and

[cmartinez127]

It's working, for now.

There is still a small issue, as you can see in the following screenshot it says " Space used: *UNKNOWN* of *UNKNOWN* ". This worked before removing the statement after "and" as you recommended.
Example of working alert: " Space used: 699.57 GB of 749.51 GB "

Screenshot: https://i.imgur.com/7V9TKr2.png

I think that now it doesn't collect data to make the Graphs.

Is there a way to solve this? Solving this is optional, I can live without it, since the alarm pops out on the dasboard that's all I need, for which I am very grateful, thanks.

[liujun0253]

You can also keep the statement after the ‘and’ pavement, so that passing past parameters can fully show the information you need. When you are familiar with the use of these expressions, adjust it to the way you want

[cmartinez127]

If I keep the statement after the "and" it no longer works. Anyways it's ok, if I ever need to collect that data in Graphs I will research and find a solution. Thank you!!