Gentle request to all community members to help me with this:
I am new to this community and same with the tool Grafana, So I am learning and integrating it with Zabbix.
I am using Grafana 10.2 and OS RHEL 8
Need to create new dashboard which will display calculated values from other dashboard.
For eg. I have one dashboard displaying CPU utilization of 10 servers, if any 2 server exceed 80% of utilization changing color to Red. I want the main dashboard to dynamically change the color to yellow and when I click that color or name of the main dashboard it will redirect to the other dashboard.
Can any one please help me how can I achieve this, and also share some links, doc or dashboards for reference.
Thanks in advance, Cheers.
Every dashboard based on a data table whatever.
Before you create the second dashboard/drilldown, you have to calculate the other values you need.
You need the new values for the thresholds to colour what ever.
I would do all these calculations in the starting table. So you can use one table for all the drilldowns.
Best way is to do this, is directly in the query. Or you use transformations.
You can use items calculated in zabbix, it will be even easier to perform this calculation. You can put all the hosts that need to be calculated in a hostgroup and then use aggredate, you would have a host only for these calculated items, later you query it through Gratana.
Hi @bernardolankheet Many Thanks for you reply.
Do you have some sample dashboard which you can share, that will help me to clarify further.
Will understand more on aggregate function on zabbix, as I am new to this zabbix as well.
Grafana query is the below image which you want to convey or anything else apart from this:
Do you have some other link which will help me to understand further, in more simpler way, that would be grateful.
Hi @joachimschiewek Many Thanks for your reply.
Will work on query and transformation part in grafana, hope that will help me.
Do you have some dashboard example which you can share from grafana dashboard. That will put more light to your explanation. Whichever data-source is fine with me, will implement and understand it better.
Hi @ghilsthomas, and welcome to the community.
Well… I advise you to always keep your business logic on the datasource’s side (zabbix in your case) and use grafana only to visualize. With that said, let’s see what you could do:
First, create a “Dummy host” for “Host Aggregated Alerts” on Zabbix, in this host, create a trigger using a logic of dependencies between the separated triggers from the hosts, here’s some docs for ya:
With the “Aggregated Dependent” Trigger in place, I would have a “Zabbix Problems” panel on my grafana side, filtered by this “Dummy host”. It’ll be clear with no lines if everything’s OK, or I’ll got my trigger popping up if I got some problem.
I know that stats and big numbers with color are cool… But my take on monitoring is more to the practical fast acting side when talking about incidents and alerts. If everything’s fine, I got nothing on my screen. If my screen got something on it, I should give it some attention.
Hope it helps.