But I could not make the package, because they’ve changed the name of phantomjs. I had phantomjs, and the script wanted phantojs-prebuild. I got it working by removing phantomjs from grafanas package.json. It’s not pretty but is works. I have phantomjs on my path.
This is for a previous release of phantomjs and unfortunately, build that way, the resulting binary has more dependency that it should and you won’t have the possibility to move it to an other distro has easily as you should. If you are interested, you can see the way they build the last version, v2.1.1, (see deploy/ on the v2.1.1 tag).
Recent versions (I think >=v3) of grafana will need a v2.1.1 binary to build cleanly.
The change in naming, is not for phantomjs binary but for the npm module that deploys phantomjs binary. It was named phantomjs in the beginning but since this was confusing, they changed the name to phantomjs-prebuild. In fact, this is the recent version of this module that will require a v2.1.1 phantomjs binary.
There are cleaner way to build grafana and deb packages can be build to with not so much trouble I you need them. If you are interested, you can read detailed steps on how to do it here. Alternately, you can simply use this .deb this PuckStar tested it on Pi3 and reported it is working fine.
Im trying to get influx and grafan running on my pi3 with hass.
Do you run influx and grafana on the same rpi3 with hass?
How many entitys do you have and how large has your influxdb get and for how many days have you run it?
I have 100 entitys and my home_assistant_v2.db is 200mb after only 1v i think I have to by a largert ssd card
Would you please summarize the steps you took to install this, it would help alot
Yes I run it on the same pi3 with HASS.
About how large it gets I’m trying to find that out myself as well
I found this info but am not sure what is means: http://imageshack.com/i/pnWMrohUp
note that your home_assistant.db is not the Influx db and that you can put a retention on that as well:
put this in your hass config:
I’m trying to find something other than InfluxDB & Grafana.
Con’s for InfluxDB:
updating old data is difficult. That is understandable because it’s meant to be time-series insert-only database. The problem is and will be that HA device names may change and I as user cannot do anything about it. Then if you have years worth of data, it can suddenly change and you have to insert the old data again. It is obvious it is not working.
Influxdb does not support joins anymore, so you cannot join data. Only one table at one query. You cannot create another mapping table from preventing the device name change problem. You cannot actually do anything because the lack of complex multi-table queries.
no functions yet to take advantage of time series data, meaning fetch data of previous and next rows.
For anyone who comes across this thread via Google - I have published a fairly detailed guide about how to set up Grafana and InfluxDB on a Raspberry Pi with HA installed on my blog. The guide is in Swedish, sadly, but Google Translate should work well. https://snillevilla.se/snygg-statistik-i-home-assistant-med-grafana-och-influxdb/
Second question : I installed succesfully InfluxDB & Grafana from add-on repository. It works fine on local network, but Grafana failed to access to Influxdb when I’m trying to access from ouside my local network (I can access to Grafana webpage and HASS webpage from ouside). Any clue ?
I think you can use iframe but you also need to expose your grafana via SSL since a ssl/tls protected page cannot iframe a non SSL/tls page.
If you have a reverse proxy (nginx, caddy etc) then set up as different subdomain etc.
I’m trying to do it but it’s not really working. Anyway, you can change environment variables of the grafana server using some options when launching the addon. For example, I’m using caddy proxy and the Infludb and Grafana addon, so, in the option you can write:
I’m sure the that the grafana addon will use these variables, but something is missing because it says:
t=2017-11-17T15:49:20+0100 lvl=eror msg=“Fail to start server” logger=server error=“open /ssl/caddy/acme/acme-v01.api.letsencrypt.org/sites/xxxx.duckdns.org/xxxx.duckdns.org.crt: permission denied”