Sorry, forgot to upload that (along with docker build files), now it’s there. it basically adds few missing pieces to get it working, such as missing plugin for grafana container and docker-ce to telegraf container.
Its great! Its work! You are genius!
But I have question. Windows docker telegraf says:
telegraf | 2019-09-26T07:30:00Z E! [inputs.disk] Error in plugin: error getting disk usage info: open /host/proc/self/mounts: no such file or directory
telegraf | 2019-09-26T07:30:00Z E! [inputs.system] Error in plugin: open /host/proc/loadavg: no such file or directory
telegraf | 2019-09-26T07:30:00Z E! [inputs.net] Error in plugin: error getting net io info: open /host/proc/net/dev: no such file or directory
telegraf | 2019-09-26T07:30:00Z E! [inputs.diskio] Error in plugin: error getting disk io info: open /host/proc/diskstats: no such file or directory
Is its normal, becouse windows has no /procs ? Or something wrong?
1 Like
I’m developing exactly that, a storj exporter for prometheus that exposes storage node api data. I have it running but needs more work. Will share my GitHub repo here soon.
That would be great! According to the last townhall, the debug output already exports everything in a prometheus compatible format but that’s a different topic.
Here it is Prometheus Storj-Exporter
@kevink Do you have more details about the debug output you mentioned please?
1 Like
Looks amazing, I’ll try it later.
They said that you can enable the debug output and point prometheus directly at it. However I did not find any information on their wiki or the forum on how to enable the debug output. It was just a configuration in config.yaml but I can’t remember (last time this configuration was mentioned was on rocketchat which doesn’t exist anymore). Maybe someone else knows it or we’ll have to open a new thread.
I found it interesting that they already have native prometheus exporting. It needs documentation. It’ll likely go to influxDB just as easily. I logged some historic debug info from rocketchat in some of the issues on my github, here:
and here:
those telegraf collectors are for ubuntu. So, on a windows system they will need to be changed for their equivalents. The full documentation on telegraf is where i’d start. Or, on @turvas github above. I’m not sure if he already converted it.
hi,
these errors are expected, as conteinerized telegraf expects /host/proc to work (was originally mapped to host /proc), but int Windows host there is no /proc equivalent. one option would be run telegraf on host (my latest github page has that already) or disable those inputs (disk, system, net) in telegraf.
br,
Here’s some info about the debug output but it doesn’t look very prometheus-like to me:
Madbrain,
this should be close enough for Win 10, have a look: https://github.com/turvas/storj_telegraf_mon
1 Like
Hi, since Version 0.29.3 the Version Panel show “0” as Version…
Nope
i found it, it was my fault. It was on my 2nd node and i forgot to set the rights for telegraf for the docker
`
sudo usermod -aG docker telegraf
`
1 Like
Do you guys get GETS and PUTS info? I don’t, and am wondering if that’s when STORJ stopped providing some of that information via API.
This info?
Only thing we’re using the API for is folder size and bandwidth used if you’re on the latest release.
Yes, am wondering if that information is still there. I refactored my dash as I couldn’t get that to populate:
1 Like
looks like you’re missing the panels altogether. You can import the current version as a new dashboard, and then copy those panels over to your custom one. Very nice board, i like the style!
Always interested in a good dashboard, I have some troubles getting it to work completely…
I run my TIG-stack on a separate machine already, while my storj-node is running docker on another separate Server 2019-machine, and I can’t get all the values to populate. And, how to get the network and cpu/ram-usage to report?
I have doublechecked the database, and it seems to miss many of the values. I guess my configuration is not supported? Any pointers? 
edit: Right, found the problem - in the successrate.sh script, I point it to the node.log-file I produce, however the command provided in the script throws an error;
./successrate.sh: line 27: >=d’ /mnt/storagenode/storj/node.log: No such file or directory
Should I create an issue on github?
Thanks, and yes, I’m aware the panels aren’t there…I removed them because they weren’t populating. Do they populate for you? If so, what script is providing that data?