#hello since i updated to nabox 3 1 2

1 messages · Page 1 of 1 (latest)

scarlet harness
#

hi @sleek ridge few ideas. Can you check your nabox logs, sounds like something is awry 🙂 power dashboard works in native harvest 22.08. The Power dashboard Nodes panel uses node_labels and environment_sensor_* Prometheus metrics. Can you also check Prometheus and see if those metrics are collected? Sounds like this used to work for you so it probably is not a permissions problem

sleek ridge
#

hi @scarlet harness . i see the environment metrics in prometheus. is it possible that i have a time problem? prometheus GUI told me: Warning: Error fetching server time: Detected 437.63699984550476 seconds time difference between your browser and the server. Prometheus relies on accurate time and time drift might cause unexpected query results.

#

can you tell me where i can find the logs? or did you mean "dc logs -f --tail 10 nabox-api"

scarlet harness
#

I think @slim shard describes how to collect the logs on this page https://nabox.org/documentation/troubleshooting/ - the full logs are typically better than tailing the last ten lines. In terms of time differences, yes that can cause problems. Not sure how @slim shard has ntp setup on the vm he's building for nabox. Hopefully he chimes in on how to configure ntp.

#

It's encouraging you see the metrics in Prometheus - let's double check that you have the latest power dashboard - what is the title of your power dashboard?

sleek ridge
#

Harvest - cDOT
/
NetApp Detail: Power

#

comes with harvest 22.05

scarlet harness
#

thanks, yep that's the dashboard from 22.05. You're on 22.08 though, right? Can you try the 22.08 dashboard instead?

sleek ridge
#

what´s the name of the 22.08 power dashboard? i made an update von 22.05 to 22.08

scarlet harness
#

sounds like when you upgraded Harvest to 22.08 that nabox did not import the 22.08 dashboards? Do you see a different folder - in Grafana when you click manage dashboards do you see a different folder for 22.08 dashboards?

#

22.08 power dashboard is named ONTAP: Power

sleek ridge
#

you´re right! i only have the "netapp detail: power" but not the "ontap: power" dashboard ..... hmmmm

scarlet harness
#

you could try to reimport the dashboards

sleek ridge
#

thats all stating with "ontap"

scarlet harness
#

yep those look correct

sleek ridge
#

and where can i find the 22.08 power dashboard?

scarlet harness
#

normally in that same list, e.g.

#

can you try reimporting the dashboards - @slim shard mentioned on Slack about a month ago, the following

sleek ridge
#

after a reset ....

#

..... i have the ontap: power dashboard

scarlet harness
#

and it works showing you beautiful power metrics !!?

sleek ridge
#

the problem is the storage node & shelf infomation in this power dashboard

#

did you see them?

scarlet harness
sleek ridge
#

but i think that is my "time" problem i have on this nabox VM

#

thank´s a lot for your help. first i´ll fix the time problem

scarlet harness
#

what do you have selected at the top for variable dropdowns?

#

do you have ClusterAll selected?

#

like so

sleek ridge
#

yes all

#

when i go back a few minutes i have everything. thats why i think the wrong time is the problem

scarlet harness
#

agreed, I'll take a quick look at the log files you shared and let you know if I see anything while you sort out ntp

sleek ridge
#

👍

scarlet harness
#

log files confirm that the shelf and sensor data is collected without error, looks peachy

INF collector/zapi.go:443 > Collected Poller=dcfast1 apiD=142ms collector=Zapi:Shelf instances=1 metrics=9 parseD=0s
INF collector/zapi.go:443 > Collected Poller=dcfast1 apiD=203ms collector=Zapi:Sensor instances=142 metrics=1116 parseD=11ms
INF collector/zapi.go:443 > Collected Poller=dcfast2 apiD=154ms collector=Zapi:Shelf instances=1 metrics=9 parseD=0s
INF collector/zapi.go:443 > Collected Poller=dcfast2 apiD=213ms collector=Zapi:Sensor instances=156 metrics=1226 parseD=9ms
slim shard
#

btw you can get rid of the old Dashboards "NetApp xxx"

#

I'll try to add a big red warning when NAbox time doesn't seem to be in sync with browser time, like the one you get on /prometheus

scarlet harness
#

seem to recall a customer on Slack having the same issue, perhaps you could update the troubleshooting section to address how to fix?

slim shard
#

The time difference or harvest dashboard reset ?

sleek ridge
#

i´m realy sorry! my vmware admin told me that he stopped the NTP service in our test environment.
service is up and running, chrony is in sync with our NTP servers and the power dashboard shows all metrics again.
have to do some adjustments on my own dashboards (temperature metrics), but now i´m happy with harvest 22.08.

greetings from austria!

slim shard
#

Normally, NAbox shouldn't use Chrony but vmware tools sync, you probably set it up yourself ?

scarlet harness