We are running Grafana version 6.1.6, running on the Linux server. We are pulling data wmi exporter, which exposes a Prometheus running on an open shift.
on dashboard 10 graph panel are there out of three panels is time out
can anyone help me on this issue
Thank you
What’s the actual query in any of the panels showing “timeout”?
What does the Query Inspector show you for any of those panels?
Can you run the query manually (I’m not familiar with Prometheus, so I
don’t know how feasible that is) to see how long it takes?
What’s the timescale for the panels?
Does the problem remain if you significantly reduce the timescale for the
graph?
Antony.
What’s the actual query in any of the panels showing “timeout”?
Calculating disk usage of group of servers
What does the Query Inspector show you for any of those panels?
Can you run the query manually (I’m not familiar with Prometheus, so I
don’t know how feasible that is) to see how long it takes?
What’s the timescale for the panels?
time scale for last 5hours
Does the problem remain if you significantly reduce the timescale for the
graph?
when i reduce the time scale to last 30 minutes it’s working
- What’s the actual query in any of the panels showing “timeout”?
Calculating disk usage of group of servers
No, when I said “what’s the actual query”, I meant show us the text of the
query, as shown in the edit window of the panel when you toggle to raw edit
mode.
In other words, show us what query Grafana is sending to Prometheus and asking
it to calculate and return the result of.
- What does the Query Inspector show you for any of those panels?
Answer?
- Can you run the query manually (I’m not familiar with Prometheus, so I
don’t know how feasible that is) to see how long it takes?
Answer?
- What’s the timescale for the panels?
time scale for last 5hours
And what resolution - how many data points are contained in 5 hours?
I wouldn’t expect a disk usage check to be sampled much more than every 5
minutes, so I’m guessing something like 60 data points (which is very modest)?
- Does the problem remain if you significantly reduce the timescale for the
graph?
when i reduce the time scale to last 30 minutes it’s working
Okay, so that tells us that the query itself is okay, but it’s taking too long
with the number of data points to be processed.
Tell us more about your back-end data store - how many data points do you have
in a 5 hour time period, and how long does a manual query take when you ask
for this data?
Antony.
