Export each Uptime (24-hour / 30-day / 1year) for prometheus endpoint

[mabed-fr]

⚠️ Please verify that this feature request has NOT been suggested before.

• I checked and didn't find similar feature request

🏷️ Feature Request Type

Other

🔖 Feature description

Export as metric each Uptime %

✔️ Solution

Export as metric each Uptime %

Ideas for name

monitor_24_hours_uptime_percent
monitor_30_days_uptime_percent

❓ Alternatives

No response

📝 Additional Context

Export as metric each Uptime %

[CommanderStorm]

• Can this not already be calculated from monitor_status via aggregation similar to the one proposed in How to view "last outage" when it happened >1 wk ago? #2982?
• What would be the advantage of adding this metric? (What kind of dashboard would you like to build?)

[mabed-fr]

Hello,

I think the data is already there:

The advantage would be to have a complete dashboard in grafana to complete this one

This would avoid calculating the value again since uptime-kuma is already there

I intend to share this dashboard once complete and reliable

[CommanderStorm]

I think the data is already there

Currently, this calculation is expensive.
See #2750 why I think this should be pushed a bit down the line.

[mabed-fr]

I am aware of #2750 and the cost for each monitor.

However, since the values ​​are already calculated, is it possible to expose them to prometheus? (24 and 30days at first?)

[CommanderStorm]

They are not already calculated (it is done on-demand afaik)
⇒ Would add additional load, which I would like to avoid until the performance problems of the current implementation are resolved.

Especially on slower disks (e.g. Rasperry Pi running of the SD-Card), this would be very noticeable.
(that is the point of the smarter aggregation proposed in #2750)
This maybe could also be overcome by the proposals in #3286, but until now, this has not yielded in an engineer testing the claims made in said issue. (If you'd like to get this feature sooner than later... 😉)

[mabed-fr]

If the value of uptime is not already calculated:/ then I understand your point of view.

[mabed-fr]

Whith the 2.0 , answer change ?

[CommanderStorm]

with v2, the data is stored in an aggregated format
=> calculating these metrics would not be expensive any more.

I have not looked at if they are included in v2.
If they are not, feel free to drop us a PR adding these as metrics..

[daveshm]

also looking forward to this feature, these metrics will be great in prometheus. looks like someone is working on a PR here: #5506