Which is the best way to monitor disk IOPS on storj node?

I have multiple nodes on different public IPs that run on same physical disk, I would like to understand if this disk is stressed.
What is the best tool to monitor this behaviour making some graphs too ?

This is against ToS so please stop doing this.

Outside of your OSes tools like iotop have a look at this one: https://www.netdata.cloud/


If you’re using ZFS: “zpool iostat” gives great counts for IO and overall bandwidth. If you’re running Linux then the iowait (wa) stat in “top” is a general indication of slowed storage (especially if you only have one disk).

In Windows general HDD activity is in Task Manager… and if you drill into Resource Monitor you can see what node (what process) is responsible for most of the load (as well as watch the Disk Queue length to see how far the disk is running behind).

1 Like

Tell me you know unvetted-nodes-get-5%-of-traffic… without telling me you know unvetted-nodes-get-5%-of-traffic :wink:

Storj decided to offer all that ingress to new nodes… can’t blame SNOs for accepting it… :money_mouth_face:

glances --diskio-iops can be used to observe the current IOPS in a terminal or in a browser, but its usefulness for monitoring Storj IOPS over time is limited.