TrueCommand resets to first-time setup continuously

[cgoebel]

I am building a TrueNAS test environment in order to evaluate the product, and I just got TrueCommand up and running last week. We are using the container image and running it on a Kubernetes node, rather than the cloud version.

The first time I accessed the TrueCommand instance it prompted me to create an admin account as expected. I created one and added our existing TrueNAS Scale system via the API key. However after closing the UI page and going back a couple hours later, rather than giving me a login prompt, it prompted me to create an admin account all over again. After doing so and logging into the system, the previously configured TrueNAS Scale instance was nowhere to be seen.

The same thing has happened over and over ever since we got the TrueCommand container running. Every few hours (I have tried to correlate it to anything specific such as length of time idle, but haven't found a strong pattern yet) the TrueCommand system will basically reset to factory default, all configurations will be lost, and I have to re-create the admin account and re-add our TrueNAS Scale test system all over again.

I could not find anything in the documentation hub or on this forum indicating anyone else had had a similar issue. If anyone knows what the issue might be, or could recommend a place to start looking into this, I would be very grateful for the assistance.

 

[sretalla]

The fact that you're running it under Kubernetes should be some kind of clue as to what's happening.

Kubernetes treats containers as ephemeral (since that's how they are documented to be used).

If (and I haven't specifically checked, so I'm just assuming) your config is saved within the container and not to some persistent storage mounted in the container, then every time the redeploy of the container happens, you'll lose that config as you start with a fresh copy of the container.

You need to find the right path and set it as persistent storage:

So I went and checked... looks like you want to use /data in the container and mount some persistent storage from there:

Installing the TrueCommand Container using Docker on Linux.

Describes the steps to install the TrueCommand container in Docker on Linux.

www.truenas.com

 

[cgoebel]

The fact that you're running it under Kubernetes should be some kind of clue as to what's happening.

Kubernetes treats containers as ephemeral (since that's how they are documented to be used).

If (and I haven't specifically checked, so I'm just assuming) your config is saved within the container and not to some persistent storage mounted in the container, then every time the redeploy of the container happens, you'll lose that config as you start with a fresh copy of the container.

You need to find the right path and set it as persistent storage:

So I went and checked... looks like you want to use /data in the container and mount some persistent storage from there:

Installing the TrueCommand Container using Docker on Linux.

Describes the steps to install the TrueCommand container in Docker on Linux.

www.truenas.com

Thank you for the detailed explanation. I spoke to my colleague who deployed the container image and they confirmed the specified :\data directory was not set to persistent. We're changing that now and I'll monitor for any further re-occurrence of the issue.