Feature Request

[csmall]

I think it would be very useful to have a way to visualize Services in a dashboard.

This would allow for more of an executive/management level overview of the overall health of a business process/application and all the supporting components.

Perhaps this is already possible using Zabbix with Grafana? I'm not very familiar with the Services functionality in Zabbix yet, but I do not see it as an option in the dashboard widgets.

[kloczek]

Monitoring -> Overview

[csmall]

Thanks, but this is not what I am looking for.

[kloczek]

So you need to describe what you are looking for.

[csmall]

So the "Services" feature of Zabbix that has all the dependencies tied together, It would be useful to have a dashboard widget to display the state of various business application "services".

So if there is a business application with 20 dependencies you could see if it is degraded or down due to a failed dependency and also display the SLA uptime for the application etc..

I guess I am looking for a dashboard widget for the Services feature.

[csmall]

Yes, you can use Grafana. http://play.grafana-zabbix.org
There is a view for IT services: https://www.zabbix.com/documentation...ng/it_services

Why there is no widget with "IT Services"? I think, because of little usability for such widget, and it can be misleading.
Why misleading? Because SLA is calculated over period of time. Let's say our important service was down for one hour - this one hour downtime will give us different SLA for different period of time. SLA for one day will be different than SLA for week or month. Widget should have bold capition with some text like "this is SLA caclulated for .....(week, month etc.)", but one will understand "month" as "month till now" and someone other as "all days of month" (even if it's future days).

In my opinion, the first changes should be in:

1) calendar management in IT services - something like in "maintenance periods". Now it is not possible to define a calendar and connect it to different IT services, you have to go to each IT service and set the calendar individually.

2) service times - now there is "Uptime" "Downtime" and "One-time downtime". There should also be "One-time uptime" - Let's say we have a calendar from Monday to Friday from 7:00 to 18:00, but in this Saturday (and only this) will be additional day work form 9:00 to 15:00. So how we can easily set this one time uptime? We can't! We me must build a new calendar week with Saturday included as normal workday, and set "one-time downtime" on all others Saturday's in month. This is very stupid but it is the only way.

3) Add number of outages in IT service view. Many outages (even without broken SLA) can mean poor quality of service (as example, server drop connection for short period of time but very often, as result you get lags). Now we have only SLA and don't know if it was one outage or more - we must search for problems in problem view.

Great points. I would also like to see sort of a healthy vs degraded marking. The service was up 80% of the time and degraded due to a failed dependent system/service that was HA and wasn't truly down 15% of the time and down 5% of the time.