Telegraf with timescaledb and materialized views

Per, Timescale db and materialized view · Issue #15590 · influxdata/telegraf · GitHub

Using telegraf 1.27.2. I am trying to setup a materialized view to down sample network tra
ffic (bytes_sent/bytes_received). I am using grafana for visualization. For some reason, I am getting very large values when I query the materialized view. My view looks like this

CREATE MATERIALIZED VIEW public.net_bytes_5min_recv WITH (timescaledb.continuous) AS 
SELECT time_bucket('5 mins',t."time") as time,tt.host,tt.interface,AVG(bytes_recv) as avg_bytes_recv 
FROM telegraf.net t,telegraf_net_tag tt WHERE (t.tag_id=tt.tag_id) group by 1,2,3 WITH NO DATA;

This creates the view.

My grafana query looks like this

WITH delta as (SELECT time,host,interface,avg_bytes_recv-lag(avg_bytes_recv,1) OVER (partition by host,interface order by time)) as bytes_recv FROM 
net_bytes_5_min_recv WHERE host in ($server) AND interface in ($interface) AND $__timeFilter("time) group by time,host,avg_bytes_recv,interface)
SELECT time,host,interface,bytes_recv from delta group by time,host,interface,bytes_recv order by 1

For some reason, I am seeing very large values for my interface. I suspect its doing a SUM() somewhere…

I was wondering if there is a good practice to do materialized views with telegraf schema.

Does any telegraf developer have a comment?

Hi,

I was curious if anyone else would comment over the long weekend, but I suspected no one would. Based on the limited information, I’m not sure this is a great question for Telegraf and really a better question elsewhere.

Without seeing the actual metrics that Telegraf is producing it is hard to say what might be at play here.