Persistent volumes are growing

I have CRDB deployed using Rancher MareketPlace (Helm Chart 4.1.200) into the K8’s cluster. It created 3 nodes and each has persistent volume attached. I did not created any DB/schema yet.
It is is running (empty) for 4.5 days and I can see that persistent volumes are constantly growing in size (they are now ~3Gi each).
Is it normal?

Hi @zeppelinux ! I haven’t used this Helm chart before, but I’d be curious to learn about what you’re seeing. What is the rate of growth? Are you saying that it started from 0gb and reached 3gb after 4 days? Is it a linear growth, or does it go up in a “spiky” way? You should be able to see in the disk capacity chart.

This might be due to timeseries data for the DB Console. Here’s an FAQ on the topic. The FAQ directly following offers ways to reduce timeseries data, if that’s what you need.

@rafiss here is the screenshot of my Longhorn volumes page:

@jesse I applied two queries for 10s & 30s and set it to 5 days (120h0m0s) will let you guys know if this is the case.

Hi Guys!

As far as i can see nothing is being deleted and it still growing, not so fast but still a bit concerning… Here is the recent Longhorn screenshot:

Let me know if you need any info/help looking at this, or may be there is something else you want me to try.

It still growing, here is the screen:

timeseries.storage.resolution_10s.ttl & timeseries.storage.resolution_30m.ttl both set to ‘120h0m0s’

Still, there is no data and even no schema that I created.

That’s pretty odd behavior! Some suggestions I can think of:

  • Can you confirm that there’s no disk usage reported on the Databases page of the DB Console? See these docs.
  • Does Rancher let you view anything else about the disk? Can you log into the node and use a command like du to view disk usage information?

Console screen looks good:

When I log in into the node and run #du -ch /
it prints only 553M total

I’m going to cross reference this post in the Longhorn forum