NTracking

[neo]

Another metric that is not what you think it should be and that is records stored. That shows both the active and inactive records combined as 1 figure - IE total records stored on disk. What people want to see is the active records.

This is something that @Shu will have to add to the metrics. We ideally would have both the active and total record counts.

Active is what the logs show when quoting is done and is in the metrics field of the log entry

Total records is what the /metrics show.

ideally both would be in the /metrics

[aatonnomicc]

yes that would make sense just updateing NTracking script now got a few gremlins to hunt down and should be good to go

[aatonnomicc]

Thats NTracking updated to use the sn_node_current_reward_wallet_balance metric.

also the readme now has a bit at the bottom for upgrading to latest version if any one is complaining send them to the readme

https://github.com/safenetforum-community/NTracking/blob/main/README.md

[aatonnomicc]

@scottefc86 @Southside

cany you guys update your NTracking script and dash board and let me know how it goes?

think I have finally fixed some of the niggles with version numbers and rewards balance being higher than discord bot thanks to @neo suggestion.

[scottefc86]

All done anything I need to keep a lookout for?

[Southside]

Before

after

Upgrading was not good - my data sources and dashboard disappeared.

I just copied all of NTracking/NTracking Dashboard at main · safenetforum-community/NTracking · GitHub and pasted it into the json box.
Which has worked OK before but not tonight…

[aatonnomicc]

did you delete the telegraf bucket and recreate it ?

[Southside]

Ah that looks a bit better now after making sure I had all Grafana windows closed and Brave restarted <— cos it needed upgraded

Yes I reinstalled Grafana Influx and Telegraf - options 2 and 3 on the menu

EDIT: Ah there is more README to RTFM…

RIght - just ignore the above and I’ll tell you in a while once the cron job runs again

[Southside]

Should I not see THREE buckets here?

[aatonnomicc]

you need to log into influx to delete the telegraf bucket that’s the screen where Grafana logs into influx to read the bucket

[Southside]

Thats better now except Im missing the earnings graph

[aatonnomicc]

that will take another update if it only has one data point I think

[Southside]

correct

[aatonnomicc]

thanks for testing and letting me know

[peca]

I just installed NTracking and it works really nice, mostly.

Where does it get these values?

This host is running 160 nodes and according to safenode-manager all are on same version and running fine.

I used to work with Zabbix, but Telegraf-Influx-Grafana is terra incognita for me. Can somebody give me a hint where to look?

[aatonnomicc]

It counts the folders in your safenode manager services folder then checks metrics ports from 13001 up to 13001 + number of folders - 1

If a node responds on metrics port +1 for node running.

If no response + 1 for stopped nodes.

[aatonnomicc]

Also if node numbers are showing nodes killed that are running it means the node timed out on the metrics request. Its a leading indicator of a machine under to much stress.

[peca]

Thanks for the info. I am going to recheck port numbers, I started nodes in batches and probably messed numbers somewhere.

[aatonnomicc]

latest update to NTracking is the happy node indicator

[riddim]

What is it built upon? I like the name - this alone makes me think about binning my setup and switching to NTracking (but my glued together stuff just works too reliable to switch horses halfway through beta =O)