Grafana refreshes -> no data

Hi everyone,

Hopefully I am in the right place for this. Currently, I am running Grafana on a virtual machine. However, I am moving things over to docker.

So, 2 instances running at the moment: one in Docker and one in the VM. Both have the same datasources.

I imported a dashboard to the new docker instance. Data is showing up, sometimes. Yes, sometimes.

See this gif for the results I’m getting: https://s7.gifyu.com/images/grafanac4cf27b54827ffa2.gif

Now, what is interesting, is that this only happens when I have several widgets. When I view just a single widget (via dropdown -> view), I do not have this experience.

I found out that this is a result for the “No data”:

Does anyone know what is happening here?

EDIT: Also interesting fact: when I hook up my InfluxDB database (the one from docker) to my VM-Grafana, I do not have this issue. So, where do I need to take a look?

Currently, I am running Grafana on a virtual machine.
However, I am moving things over to docker.

Is that on the same underlying platform, or different hardware?

So, 2 instances running at the moment: one in Docker and one in the VM.
Both have the same datasources.

I wonder whether your Dockerised version is experiencing lower CPU performance
than the VM version gets, based on the host system?

Antony.

@pooh

Thanks for your answer. I am running on ESXI.
The VM has 4 CPU cores and 1 socket. 2GB Ram.

The Docker virtual machine:
4 CPU cores and 1 socket, 32GB Ram.

I can try to tweak with the resources, but that would not explain why my dockerized InfluxDB container works perfect on the virtual machine Grafana, no?

EDIT: Also, I exprience no speed issues or whatsoever on Docker. All really fast.
EDIT2: I gave the Docker VM double the CPU resources - same behaviour. I gave the Grafana VM less resources - works fine.

@pooh could this be a timezone related issue? When I run the image on my local machine - no issues with the same influxdb datasource. Can you think along with me? :slight_smile:

  1. All data for Grafana should be held in the back-end data store in UTC.
    Please confirm.

  2. If you suspect a timezone problem, change the timezone of your Docker
    machine to match that of the VM (and/or vice versa) and see if the problem
    moves with the timezone change or stays with the architecture.

Regards,

Antony.

@pooh okay, I found the solution.

In my network (docker-network), I was running another host which had the same name as the InfluxDB container. Not sure how that is possible, but it was load-balancing between the two containers. Solved!