Server Uptime Percentage Calculation Using system_uptime Metric

I am attempting to calculate the overall uptime percentage of a server using the system_uptime metric, incorporating both the increase and reset functions. However, the resulting value appears to be lower than expected. Could you please assist me in configuring the appropriate query and setting up the panel correctly?
Grafana version: V12.1.0

query :

(sum(increase(system_uptime{instance=“$instance:9273”}[30d])) / (30 * 24 * 60 * 60)) * 100

Hi, logically speaking, the query is fine. I tested it on some of my host that restarted two days ago and the percentage was close to 99%. Can you share why do you think the value is lower than expected?

can you please run for a longer time, 180 or 360 days. ?

My retention has max 90 days, so unfortunately no

how about for the 90 days, it was working ?

Yeah, it worked for me - why do you say it didn’t?