InfluxDB relay High Availability

Hi! I am working on a simple InfluxDB High Availability setup using Relay. Writing and querying data works fine however, changing database configuration became very inconvenient. Essentially, now I have to execute each command (e.g. CREATE USER …) twice - once for each InfluxDB instance.

In order to solve this, I need to make sure that once any configuration changed on one node, it is reflected on another one. From what I’ve learned from the doc, I understand that InfluxDB stores data and meta configuration separately in (“data” and “meta” directories respectively). My idea was to add some NFS shared by both nodes and store meta there, this way no matter on which instance configuration has changed, it will be visible from another instance since they have a common “meta” directory.

The questions are: Does it really work this way? Is it safe to use it like this?

Thanks in advance!

Did you get any answer for this? I am looking for the same.


No, I didn’t unfortunately. For me it doesn’t matter so much now, since we a moving to Prometheus.