On Mon 16 Nov 2026 at 12:24:27 (+0100), Vincent Lefevre wrote:
> On 2026-11-13 13:40:30 -0600, David Wright wrote:
> > On Fri 13 Nov 2026 at 17:55:06 (+0100), Vincent Lefevre wrote:
> > > In addition to iperf, only bash-completion has a match.
> > > I haven't deleted any package. There have been 2 iperf
> > > upgrades, though. And it has been created during the
> > > first one.
>
> Indeed, the iperf user was created when iperf 2.2.1+dfsg-3 got
> installed, and this was not expected.
>
> > In the package iperf_2.2.2+dfsg2-1_amd64.deb, the contents of the file
> > /usr/lib/sysusers.d/iperf.conf looks as though it might be relevant:
> >
> > u! _iperf - "iperf daemon"
> >
> > Same for iperf3 (just add a 3.)
> >
> > Perhaps iperf{,3} does it itself, or perhaps systemd?
>
> That's _iperf, not iperf. But the iperf.service file still had
> User=iperf. Then I reported a bug:
>
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1153939
>
> and the maintainer fixed various things (and iperf has been split
> into iperf-client and iperf-server). In particular, the iperf user
> has been removed by the iperf upgrade, so that now, everything is
> consistent on my machines.
It does seem that systemd expected there to be an iperf user,
judging by the iperf.service file in certain recent versions:
[Service]
Type=simple
Restart=always
RestartSec=15
User=iperf
ExecStart=/usr/bin/iperf --server --interval 0 --ipv6_domain
SuccessExitStatus=1
but suspecting ‘u! _iperf - "iperf daemon"’ was guesswork on my part
as I can't parse such gobbledegook. I don't what does parse it, or
how systemd would deal with a missing user.
I don't think trixie's iperf knows about systemd, so I can only
guess that one of the iperf_2.2.2+dfsg2-?_amd64.deb packages
was responsible, as you said. Now that iperf_2.2.2+dfsg2-5_all.deb
exists, which deletes any iperf and _iperf users, I guess that other
people won't be affected by this glitch if they weren't active
around 12–14 November.
Cheers,
David.