Triage report (2020-05-04, Monday triage)

Paride Legovini paride.legovini at
Tue May 5 10:17:52 UTC 2020

# Triage

LP: #1875926 - *(Confirmed)     [net-snmp]       - snmpd upgrade
(Bionic->Focal) changes Debian-snmp UID/GID

Marked as Triaged for Bionic as the bug is well understood now (thanks

LP: #1802628 - *(Confirmed)     [samba]          - winbind name
resolution fails when remote host has multiple IPs

2018 bug, apparently still an issue in Focal. The corresponding Debian
bug is from 2016, and while it has been marked as Done the problem
wasn't claimed as resolved. Difficult to act on this one.

LP: #1876622 - (New)            [samba]          - Can't upgrade
samba-common-bin: undefined symbol: smb_strtoul

This was double-triaged by me and Andreas :) My fault keeping the
browser tab open too long before hitting "Post Comment", and I didn't
see Andrea's reply. Anyway attempted to install samba from source,
compiling it, and most likely had a half-broken mix of his own compiled
version and filed from the .deb.

The bug is Incomplete for the moment, but can be considered Invalid.

LP: #1876564 - (New)            [dovecot]        - mail-stack-delivery
has missing packages on focal

Some users are confused by the fact that the mail-stack-delivery
metapackage in Focal does not behave like in Bionic, that is: does not
setup the full mail stack. In fact, it's now a basically empty
metapacakge. The change was not very easy to spot and we should probably
have dropped the metapackage entirely in Focal, as some d/changelog
entries suggest:

  - carry mail-stack-delivery as empty transitional package

    (can be dropped >20.04)

Moreover, the server guide still mentions the metapackage as a way to
setup a mail stack.

I subscribed the team as we want to improve on the current situation I
think, at least to upgrade the server guide.

LP: #1875771 - *(New)           [freeipmi]       - ipmi_locate segfault
on Focal

That's it: impi-locale segfaults on Focal on specific hardware. The bug
is from Jeff Lane (Enablement). I think the only way forward is to get
access to the affected hardware. What do you think?

Marked server-triage-discuss

LP: #1876320 [openssh] - Port parameter sshd_config is 22 AND whatever
you specify

An "Include" line in sshd_config causes "Port 22" to be added to the
sshd configuration, even if a different port is explicitly configured
after the Include line. This causes sshd to listen on both ports.

I think this is an upstream bug, and asked the bug submitter if he's
willing to report it upstream. If he does not, I think we should do it.
I subscribed the team.


More information about the ubuntu-server mailing list