Graph from influuxdb


I’m a new user on Grafana and Influxdb, I’m collecting traffic from ethernet interfaces (InOctets/OutOctets) with Telegraf and storing at InfluxDB.

Influxdb data saved…

> select * from system
name: system
time                InOctets   InUcastPkts OperStatus OutOctets  OutUcastPkts    agent_host host   sysUpTime
----                --------   ----------- ---------- ---------  ------------    ---------- ----   ---------
1492215750000000000 2167862960 1198272519  1          89039439   5.18488755e+08  se-csc1a   xanadu 1172388658
1492215755000000000 4231935343 1199833560  1          233425338  5.19276498e+08  se-csc1a   xanadu 1172389150
1492215760000000000 2062913075 1201442774  1          379415053  5.20096832e+08  se-csc1a   xanadu 1172389650

When I configure Grafana dashboard with graph from series above, the result is not correct in compare with same graph from MRTG system that I have use today.

following pictures:

How to graph Grafana as MRTG?



Sorry very hard to say. Is the data in your two graph collected in the same way, are they queries the same way? What doe sthe raw data look like?

Impossible to answer your question as it is now.You need MUUUUCH more info. Exactly what your collecting, what your previous query and data looks like what your new data looks like and how it is collected, what does your new data mean, (network data can be collected in very differently).

I would ask the InfluxDB guys or the ones who wrote your metric collector (the thing that is recording the data)

Check the time intervals being presented. Is every point on the MRTG graph the same as every point on the Grafana graph? As I have been going through adding things to Grafana that we have/had elsewhere, Grafana/InfluxDB has had more granular representation than the other systems. I have learned to use the ‘Group by time interval’ box further down the page.

For example, some of my graphs in other systems are displaying a 1 minute or 5 minute datapoint. My Grafana/InfluxDB data is using 10 second intervals by default. If I set the ‘Group by time interval’ value to be the same as the datapoints in the other systems, then the graphs look fairly similar. A mean value over 10 seconds will be different than a mean value over 5 minutes, based on the spikes in your graph.

1 Like

Hi Guys,

After read your answers, I compared mrtg config.

I adjusted the intervals and scale (math), now graph is OK.

Thanks for help…

Follow my graph and query.

So, I have a new question… :slight_smile:

Field’s alias show always the metric name, exemple: uplink_traffic.IN, How show only “IN” at graph?

Tks again.

The metric name is and field alias is always returns as the series name by InfluxDB. If you want to change that you the ALIAS BY field and alias patterns (see help of available alias patterns using the help topic below the query editor). for example $col top use the column name only as the series name.


Thanks. My Graph it woriking…