Telegraf Monitoring of PDU with REDFISH

Telegraf
1

We have ServerTech PDUs that support REDFISH, and the output of REDFISH is a lot easier to assess than the SNMP options. It appears that the REDFISH plugin may only be intended to work with compute systems?

This plugin gathers metrics and status information of server hardware with enabled DMTF’s Redfish support.

Is there any way to work around this (This plugin or a different plugin)

If I do this:

curl -kfsS -u “${AuthString}” “https://${Ip}/redfish/v1/PowerEquipment/RackPDUs/1/Mains/I1” | jq ‘.“PolyPhaseCurrentAmps”’

I get this:

{
“Line1”: {
“Reading”: 5.287,
“CrestFactor”: 1.481,
“THDPercent”: 6.3000000000000012
},
“Line2”: {
“Reading”: 5.579,
“CrestFactor”: 1.498,
“THDPercent”: 5.9
},
“Line3”: {
“Reading”: 6.468,
“CrestFactor”: 1.47,
“THDPercent”: 4.3
}
}

I would love to be able to process that with telegraf

2

I think I should be able to do something like this?

[[inputs.exec]]
commands = [‘curl -kfsS -u “LogonName:LogonPassword” “https://10.0.0.1/redfish/v1/PowerEquipment/RackPDUs/1/Mains/I1”’\]
interval = “10s”
data_format = “json_v2”
json_query = ‘.’
key_tags = [“PolyPhaseCurrentAmps”]
name_override = “redfish_pdu_metrics”

Am I even on the right track here?

3

I have never dealt with redfish before, but based on the command you have if that doesn’t work it would be pretty easy to put that inside of a python script and just have inputs.exec run that python script. It might give you a little bit better parsing options (depending what you need the output to look like).

4

Well, I guess this is not a thing (thank you AI for the delay :laughing: )

data_format = “json_v2“

Looks like it must be:

data_format = “json“

Now I get everything that is in the returned JSON, or if I can use metric filtering with exec to cherry pick what I want.