Netdata configuration

Configuration
1

I’m struggeling to understand the data_group and element in the new netdata component (HA 0.71). For example I tried:

system.net:
  element: in

and

system.net:
  element: received

But I get and error; Sensor is not available: system.net

Using system.ipv4 results in an error; netdata: Error on device update!

2

Same here.
Thought this would work:

hddtemp_local.disks_temp:
  element: sdd

But nope :confused:

3

Maybe @fabaff can help us out on this.

4

Unfortunately are there at least two bugs:

  • Multiple entries for the same resource will not work. Doesn’t pass configuration validation.
    resources:
      cpu.cpufreq:
        element: cpu0
      cpu.cpufreq:
        element: cpu1
  • There are some deviation in the naming which I didn’t caught during the Beta. Like for hddtemp 
"hddtemp_local.temperatures":{  
   "name":"hddtemp_local.disks_temp",
   "context":"hddtemp.temperatures",
   "units":"Celsius",
   "last_updated":1528578157,
   "dimensions":{  
      "sda":{  
         "name":"sda",
         "value":45
      }
   }
},

@LarsNorgaard used hddtemp_local.disks_temp which is shown in the frontend but in the response it is
hddtemp_local.temperatures. Thus the sensor is not working.

This seems also be the case for network metrics.

"system.ipv4":{  
   "name":"system.ipv4",
   "context":"system.ipv4",
   "units":"kilobits/s",
   "last_updated":1528578181,
   "dimensions":{  
      "InOctets":{  
         "name":"received",
         "value":2.4819415
      },
      "OutOctets":{  
         "name":"sent",
         "value":-2.4589827
      }
   }
},

I will try to fix it in the next couple of days and include it in a bugfix release if possible. I’m sorry for the inconvenience.

As an ugly work-around access http://[IP_NETDATA_INSTANCE]:19999/api/v1/allmetrics?format=json with your browser then you will get the raw data that is used. Search for the data_group and check the names of the element.

5

I opened an issue to track this:

6

Before I had

  - platform: netdata
    host: 192.168.1.239
    name: Pi2_Raspberry
    resources:
      - disk_free
      - memory_free
      - system_load

what is the correct configuration now, I do not understand it from the docs

7

As @fabaff said, go to the API json link: [netdataIP]:19999/api/v1/allmetrics?format=json and search for the data you want to show.

Ie. search for system.cpu on that api json link and you will see the element you need is called system.

If you want to show available ram you’ll have to search for mem.available and see that the element you need in your yaml is called MemAvailable.

  - platform: netdata
    host: 192.168.1.10
    resources:
      system.load:
        element: load15
      system.cpu:
        element: system
      hddtemp_local.temperatures:
        element: sdd
      mem.available:
        element: MemAvailable
8

thanks, you know which is the command for disk availablle (free amount of HDD?)

This?

  disk_space._run:
    element: avail
9

Yes, but I would use disk_space._ instead. That one should give you the available disk space for the root drive.

Note though that it only shows the size of the drive and not how much space is free.

10

Well I need to know the free space, that’s the whole point for the need of a sensor

11

Docs update is on the way

12

Im not sure if that doc update has made it or not I just seen these docs for the first time. But its not very clear. I can see netdata json, Id like to monitor cpu/memory/network/disk can someone shoot an example out? its an ubuntu 16.04 box…

thanks

13

sorry to revive an old thread, but I am battling to get this component configured 100% correctly.

Can I just ask for clarification that the element is marked “A” in the pic or is it “B”?

image

Thanks for any help…