1

Running Grafana 5.4.2-1 + InfluxDB 1.7.4-1 + Telegraf 1.9.4-1 on CentOS 7.6.

I used to run a monitoring stack on collectd -> graphite -> grafana, where say, a single metrics (CPU Idle) graph can show multiple polled hosts into the same dashboard graph.

Right now, I'm unable to replicate such behaviour with telegraf -> influxdb -> grafana. Every dashboard I've checked on grafana.com provide a drop-down menu to select which host to display, but none offer the capability to show multiple hosts for the same metrics.

Is it that InfluxDB is structurally unable to do so, or that I am missing something (and if so, what am I missing ?).

Thanks

1 Answers1

0

Well, I've received no answer in about a week.
The main reason I was switching from collectd->graphite->grafana to TIG is that somehow I had breakage all over the place with graphite/django, did not want to investigate and felt it was time I'd switch to something more "modern" and less prone to depency hell.

Since nobody replied, I investigated my ansible playbooks, tweaked them, and went to an hybrid betwen collectd->graphite->grafana: telegraf->graphite->grafana.

Waddayaknow.. it works just fine