MariaDB or InfluxDB on Synology

Hello together,

I want to make next step and collect some sensors values to show it as graphs.
What I have found, the Grafana is for now the best way for that. Also should InfluxDB be better for that as MariaDB, right?
Now my question. I have my HA on Synology in a docker. What I see, I can directly by DSM (Synology) install the MariaDB (the InfluxDB is not available). Would be it not the better way to install DB directly by DSM as inside HA? I’m asking because of the future when I perhaps need a databank to don’t install a second one?
What would You do?

Best,
kamil

If you dont have influx available, only mariadb, then that would be fine too.

But grafana + influxdb is the way to go :slight_smile:

hmm, I have found, that it is possible to install the InfluxDB as container in docker. I think it is not so easy as MariaDB as package but if You say Influx is better as Maria…

But what is better. To install the database inside HA or inside Docker/Synology?

HA addon comes with the easy configuration part. Just press install, follow the instructions and you’re done. The community will maintain the addon, so its good to start there.

It all depends on your level of expertice. If you have good docker understanding and know how to handle it from there, then I’d suggest you try that.

I run influxdb on a seperate docker container, because I run home assistant core (no supervisor, thus no addons)

@kamcio

MariaDB is essentially MySQL- so you will receive a speed boost on the history pages over the default sqllite file. So if you want all data from within HA, then Maria DB is for you (maybe I prefer PostgreSQL even over Maria DB)

InfluxDB is a time series database that you can push sensor data and other entities into from Home Assistant, but serves no purpose to Home Assistant itself. It’s so you can then create dashboards and such using Grafana.

Hope this clarifies a bit and you can make a better choice.

Heck you can even run both, MariaDB and InfluxDB.

In my setup I only have 2 sensors that I want to have history access too from within HA (that I write to local sqlite db. Not too intensive)

All other domains (lights, all sensors, alarm, climate etc etc) i write to influxdb with no purge date. It is build for that kind of data, and I generate grafana graphs with data i want to display back to HA as png.

2 Likes

@Vasco
many thanks for Your explanation.
If I do something, I want to make it from beginning correct. I don’t like to install something and after some time deinstall it. Because of that I ask You.

Really? I’m still not decided which way should I take because how I understand, nothing is really advised against. But I think, I will stay by Influx inside HA + grafana…

If you want a long term solution, then I would advise you to install influxDB to record all domains worth recording (alarm, climate, lights, sensors, boolean?) etc. Then use grafana to turn that data into graphs and publish those graphs to HA.

Decide whether you need history data from HA directly, (using automations based on that data)? If so, and it’s a lot then I advise to install MariaDB too and log everything you need for your automations etc there. If its a few sensors only, then go with sqlite. If you need none, then disable recorder for HA and only use influxdb + grafana.

@Vasco
I think, I will in the future record many domains, especially the sockets to see how many electricity use each device.

For now I have only two automations in HA… but I will keep that in mind to install also MariaDB when it will be more.

Thank You Vasco!

1 Like