#ONTAP:SVM Data Not Showing
1 messages · Page 1 of 1 (latest)
@covert elbow It is supported. Can you share Harvest logs @ https://upload.nabox.org/pade-yity-kabi
Uploaded
@covert elbow I don’t see anything in the logs that indicates why the SVM dashboard is empty. If we focus on one panel in the SVM dashboard ,SVM average latency. Are you saying this panel is empty? If so, could you check whether there is any data for the metrics svm_vol_read_latency and svm_vol_write_latency in the time-series database? Do you see performance data in volume dashboard?
All of the panels mostly at the top for anything performance related just says “no data”.
How about volume dashboard? Does it have performance data in highlights section? If you run above metrics queries in time series db. Do they return anything?
volume dashboard I see the performance data.
The timing of this seems to have happened AFTER I ran out of space on my /data directory and expanded the drive and restarted machine.
But maybe that is nothing and just coincidence.
hi @covert elbow can you check in VictoriaMetrics UI for the metrics that Rahul asked about? svm_vol_read_latency and svm_vol_write_latency
WARNING: No match! This query hasn't selected any time series from database.
For both
Is that what you needed?
yes, that is unexpected
WARNING: No match! This query hasn't selected any time series from database. Either the requested metrics are missing in the database, or there is a typo in series selector.
ok so VM does not have the metric, that's why the dashboard is blank. Now, why doesn't VM have the metric. The poller should be exporting it so why isn't it. Let me check your templates more closely - we noticed that you have made several modifications to them so maybe something when awry there
Hmmm, I haven't made any changes to templates. I pretty much use this "as is"..
looks like you have a restperf custom.yaml for volume and since the missing metrics come form the volume perfomance collector it makes sense to focus there. Can you try disabling you custom.yaml, restarting the containers (dc down; dc up -d), and then checking if the metrics are available after 5 minutes or so?
Let me check that out now and get back soon.
There are a couple of locations but since I use REST I would assume it's this location - /etc/nabox/harvest/active/restperf: 9.12.0 custom.yaml
it is etc/nabox/harvest/user/restperf/custom.yaml since active is created at runtime by nabox - try renaming or mv this file
That looks better so do I just leave it disabled?
It is maybe an incorrect edit I made in the file or I just don't need it?
maybe you originally added it because you wanted to collect metrics for flexgroup constituents? If so, we can reenable that but for the correct template
What's the right template? Although my FlexGroup report does have data currently.
Based on this I thought the disabled custom.yaml template in the location we just disabled was correct - https://github.com/NetApp/harvest/discussions/3905
Awww it was a little different - mkdir -p /etc/nabox/harvest/user/rest
discussion 3905 mentions keyperf and zapiperf, not restperf. We removed your custom restperf template
Well I have a custom.yaml at this location /etc/nabox/harvest/user/rest and my Flexgroup data is mostly showing.
I created that custom.yaml long ago
yes, that will ensure the config-related flexgroup information is collected but that template will not collect performance metrics, only config.
I think what happened is a long time ago, you enable flexgroups with your custom yamls by following https://github.com/NetApp/harvest/discussions/3778 Notice this mentions chaning rest and restperf, which is what you did.
But 25.11 changed that, as mentioned in the release notes and https://github.com/NetApp/harvest/discussions/3905
That is correct because there is no performance data.
make sense. To add performance data follow the directions in #3905 for "Extend keyperf volume template"
Got it, I will follow those steps.
Funny. Anytime I still go and edit the harvest.yaml file I run into the same issue. I throw KeyPerf into the file and once again nothing changes.
I figure I will just leave the performance out for now.
we can get it working! Can you restart everything again, wait 5m, grab a support bundle and upload to https://upload.nabox.org/pade-yity-kabi again?
What was different this time?
I used a ASCII editori
ah ok, so maybe tabs vs space issue in the yaml
Yep
glad you got it working
Thanks for your help as always!
you're welcome!
Why is it when I add the KeyPerf to more than one system the other one stops working and the new one doesn't report any performance either?
I think I'll upload the logs.
Uploaded
can you tell me which poller is working and which one isnt?
{"log":"time=2026-01-15T18:55:51.482Z level=ERROR source=conf.go:75 msg="unable to load config" error="yaml: line 186: found a tab character that violates indentation"\n","stream":"stdout","time":
"2026-01-15T18:55:51.482301225Z"}
As in the actual system name?
the poller name
that's the collector name, the poller name is the name listed in the Pollers section of your harvest.yml
Without saying names, the very last system in the harvest.yaml file.
you said: "Why is it when I add the KeyPerf to more than one system the other one stops working and the new one doesn't report any performance either? "
I need to know the names of the system to compare them. If you can't say the poller name, please email us at ng-harvest-files@netapp.com and tell us the poller names
Sent