On Wed, Jan 14, 2026 at 09:09:40AM +0000, Olivier Cherrier wrote:
> I tested it on a machine with 2 disks and it worked well.

Thank you for taking a look, and for the original script!

> Sure it is handy to have all disk stats on the same graph.
> We can definitively inlude it into net/munin port.

Yup, and if you click you get separate graphs! Best of both worlds :)

> BTW, I've never had issue with the current iostat_ script. But it
> requires to list the devices you want to monitor.

I ran into this issue with 'iostat_' on a 2 disk system:

	# sysctl hw.disknames
	hw.disknames=cd0:,sd0:0fcf2b034ea7b46f,sd1:07ef68266e9cd598,fd0:

	# ls -al /etc/munin/plugins/iostat_sd*
	lrwxr-xr-x  1 root  wheel  40 Jan 10 10:30 /etc/munin/plugins/iostat_sd0 -> /usr/local/libexec/munin/plugins/iostat_
	lrwxr-xr-x  1 root  wheel  40 Jan 10 10:30 /etc/munin/plugins/iostat_sd1 -> /usr/local/libexec/munin/plugins/iostat_

	# grep iostat_ /var/log/munin/munin-update.log | tail -3
	2026/01/12 11:05:07 [WARNING] Service collision: plugin iostat_sd0 on localhost (127.0.0.1:4949) reports bytes which already exists on that host.  Deleting new data.
	2026/01/12 11:10:02 [WARNING] Service collision: plugin iostat_sd0 on localhost (127.0.0.1:4949) reports xfr which already exists on that host.  Deleting new data.
	2026/01/12 11:10:02 [WARNING] Service collision: plugin iostat_sd0 on localhost (127.0.0.1:4949) reports bytes which already exists on that host.  Deleting new data.

Anyway, I'm not sure we should keep both 'iostat_' and 'iostat' around (since
they are intended do the same thing). 

Kind regards,

Job