Volume disk usage on dashboard

[lubien]

@danwetherald

This is awesome. What is the best suggested way to monitor this and receive alerts when the storage is at 90% usage for example.

We ship file system usage to our Prometheus so you could setup something like self hosted grafana alerting.

@danwetherald

Also, a lot of our PG apps are now showing “No volumes” but all 3 machines are running for example.

Are they on Apps V2? I should have added to the notes that this will work only for apps v2. We are moving towards all apps using machines so our goal is for those databases to be able to benefit from new work like this one. We will be adding a notice there soon.

@LeoAdamek

The numbers seem off for me, our pg_data volumes show 1,378MB of usage in the dashboard – but df -h reports only 480MB of usage… are volumes replicated in some way and are the replicas being counted?

It could be slightly off indeed, even df might show off numbers. Here’s an example of that on a machine of mine:

root@5683d269f3758e:/# df -h
Filesystem      Size  Used Avail Use% Mounted on
devtmpfs        2.0G     0  2.0G   0% /dev
/dev/vda        7.9G  806M  6.7G  11% /
shm             2.0G  1.1M  2.0G   1% /dev/shm
tmpfs           2.0G     0  2.0G   0% /sys/fs/cgroup
/dev/vdb        9.8G  149M  9.2G   2% /data

You can see 5683d269f3758e is one of the attached machines on the screenshot above, it says on the dashboard the usage is 709MB while df says 149M but opon checking the avail column on df you can spot the weirdness.