* [pve-devel] [PATCH pve-manager stable-8 v1 0/1] Note Potential Audit Log Message Flood in pve8to9
@ 2025-08-04 16:25 Max R. Carrara
2025-08-04 16:25 ` [pve-devel] [PATCH pve-manager stable-8 v1 1/1] pve8to9: check if unit 'systemd-journald-audit.socket' is active Max R. Carrara
0 siblings, 1 reply; 4+ messages in thread
From: Max R. Carrara @ 2025-08-04 16:25 UTC (permalink / raw)
To: pve-devel
Note Potential Audit Log Message Flood in pve8to9 - v1
======================================================
In short, emit a NOTICE message if the 'systemd-journald-audit.socket'
unit is active in pve8to9, suggesting the user that they should stop and
disable the unit to prevent audit message spam in syslogs.
Summary of Changes
------------------
Max R. Carrara (1):
pve8to9: check if unit 'systemd-journald-audit.socket' is active
PVE/CLI/pve8to9.pm | 51 ++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 51 insertions(+)
--
2.39.5
_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
^ permalink raw reply [flat|nested] 4+ messages in thread
* [pve-devel] [PATCH pve-manager stable-8 v1 1/1] pve8to9: check if unit 'systemd-journald-audit.socket' is active
2025-08-04 16:25 [pve-devel] [PATCH pve-manager stable-8 v1 0/1] Note Potential Audit Log Message Flood in pve8to9 Max R. Carrara
@ 2025-08-04 16:25 ` Max R. Carrara
2025-08-08 8:15 ` Fabian Grünbichler
0 siblings, 1 reply; 4+ messages in thread
From: Max R. Carrara @ 2025-08-04 16:25 UTC (permalink / raw)
To: pve-devel
... and display a `log_notice()` to the user if it is in order to
inform them about audit messages being logged during the upgrade,
as well as suggesting that they stop and disable the unit before
the upgrade.
In Debian Bookworm and earlier, audit messages were not logged by
default [0], whereas the unit was active by default. This was changed
in between Bookworm and Trixie; the patch that changed the default was
dropped [1], whereas the unit is now disabled.
This means that the unit will remain active during the upgrade, which
will in turn cause a lot of audit message log spam.
[0]: https://salsa.debian.org/systemd-team/systemd/-/commit/07daa6196f9c92be8a0f552b1416576e80d054dc
[1]: https://salsa.debian.org/systemd-team/systemd/-/commit/7c6ea97a1d7e438e6621c3b97ce472754fd3db43
Signed-off-by: Max R. Carrara <m.carrara@proxmox.com>
---
PVE/CLI/pve8to9.pm | 51 ++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 51 insertions(+)
diff --git a/PVE/CLI/pve8to9.pm b/PVE/CLI/pve8to9.pm
index 426fa8a6..bd300d89 100644
--- a/PVE/CLI/pve8to9.pm
+++ b/PVE/CLI/pve8to9.pm
@@ -2129,6 +2129,56 @@ sub check_legacy_sysctl_conf {
log_pass("Legacy file '$fn' exists but does not contain any settings.");
}
+sub check_journald_audit_socket {
+ my $unit = 'systemd-journald-audit.socket';
+
+ log_info("Checking whether systemd unit '$unit' is active...");
+
+ my ($res_is_enabled, $res_is_active) = (undef, undef);
+
+ eval {
+ run_command(
+ ['systemctl', 'is-enabled', "$unit"],
+ outfunc => sub {
+ ($res_is_enabled) = @_;
+ chomp $res_is_enabled;
+ },
+ noerr => 1,
+ );
+ };
+
+ eval {
+ run_command(
+ ['systemctl', 'is-active', "$unit"],
+ outfunc => sub {
+ ($res_is_active) = @_;
+ chomp $res_is_active;
+ },
+ noerr => 1,
+ );
+ };
+
+ if (!defined($res_is_enabled) || !defined($res_is_active) || $res_is_enabled eq 'not-found') {
+ log_skip("Unit '$unit' not found");
+ return undef;
+ }
+
+ my $msg =
+ "Unit '$unit' is '$res_is_enabled' and '$res_is_active' - to prevent an excessive amount of"
+ . " audit messages being logged during the upgrade, it is recommended to stop and disable"
+ . " '$unit' beforehand."
+ . "\nThis can be achieved by running the following command:"
+ . "\n\n\tsystemctl disable --now $unit\n";
+
+ if ($res_is_enabled eq 'enabled' || $res_is_active eq 'active') {
+ log_notice($msg);
+ return undef;
+ }
+
+ log_pass("Unit '$unit' is '$res_is_enabled' and '$res_is_active'");
+ return undef;
+}
+
sub check_misc {
print_header("MISCELLANEOUS CHECKS");
my $ssh_config = eval { PVE::Tools::file_get_contents('/root/.ssh/config') };
@@ -2225,6 +2275,7 @@ sub check_misc {
check_rrd_migration();
check_legacy_ipam_files();
check_legacy_sysctl_conf();
+ check_journald_audit_socket();
}
my sub colored_if {
--
2.39.5
_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [pve-devel] [PATCH pve-manager stable-8 v1 1/1] pve8to9: check if unit 'systemd-journald-audit.socket' is active
2025-08-04 16:25 ` [pve-devel] [PATCH pve-manager stable-8 v1 1/1] pve8to9: check if unit 'systemd-journald-audit.socket' is active Max R. Carrara
@ 2025-08-08 8:15 ` Fabian Grünbichler
2025-08-08 11:16 ` Max R. Carrara
0 siblings, 1 reply; 4+ messages in thread
From: Fabian Grünbichler @ 2025-08-08 8:15 UTC (permalink / raw)
To: Proxmox VE development discussion
On August 4, 2025 6:25 pm, Max R. Carrara wrote:
> ... and display a `log_notice()` to the user if it is in order to
> inform them about audit messages being logged during the upgrade,
> as well as suggesting that they stop and disable the unit before
> the upgrade.
>
> In Debian Bookworm and earlier, audit messages were not logged by
> default [0], whereas the unit was active by default. This was changed
> in between Bookworm and Trixie; the patch that changed the default was
> dropped [1], whereas the unit is now disabled.
>
> This means that the unit will remain active during the upgrade, which
> will in turn cause a lot of audit message log spam.
so this will warn about something that happens on every system(!), but
is only relevant for the duration of the upgrade and just makes the logs
more noisy?
I am not sure that tradeoff is worth it, it is already noted in the
upgrade guide, so people who read that will do it proactively anyway,
and people who notice it during the upgrade can check there as well and
see that it's a non-issue..
or maybe we could move it to `--full`?
adding too many warnings (and lines in general) makes it much easier to
miss actually relevant things..
in particular, after the upgrade is done it makes no sense to check this
anymore, since the next step will be a reboot that gets rid of the
excessive logging anyway..
>
> [0]: https://salsa.debian.org/systemd-team/systemd/-/commit/07daa6196f9c92be8a0f552b1416576e80d054dc
> [1]: https://salsa.debian.org/systemd-team/systemd/-/commit/7c6ea97a1d7e438e6621c3b97ce472754fd3db43
>
> Signed-off-by: Max R. Carrara <m.carrara@proxmox.com>
> ---
> PVE/CLI/pve8to9.pm | 51 ++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 51 insertions(+)
>
> diff --git a/PVE/CLI/pve8to9.pm b/PVE/CLI/pve8to9.pm
> index 426fa8a6..bd300d89 100644
> --- a/PVE/CLI/pve8to9.pm
> +++ b/PVE/CLI/pve8to9.pm
> @@ -2129,6 +2129,56 @@ sub check_legacy_sysctl_conf {
> log_pass("Legacy file '$fn' exists but does not contain any settings.");
> }
>
> +sub check_journald_audit_socket {
> + my $unit = 'systemd-journald-audit.socket';
> +
> + log_info("Checking whether systemd unit '$unit' is active...");
> +
> + my ($res_is_enabled, $res_is_active) = (undef, undef);
> +
> + eval {
> + run_command(
> + ['systemctl', 'is-enabled', "$unit"],
> + outfunc => sub {
> + ($res_is_enabled) = @_;
> + chomp $res_is_enabled;
> + },
> + noerr => 1,
> + );
> + };
> +
> + eval {
> + run_command(
> + ['systemctl', 'is-active', "$unit"],
> + outfunc => sub {
> + ($res_is_active) = @_;
> + chomp $res_is_active;
> + },
> + noerr => 1,
> + );
> + };
> +
> + if (!defined($res_is_enabled) || !defined($res_is_active) || $res_is_enabled eq 'not-found') {
> + log_skip("Unit '$unit' not found");
> + return undef;
> + }
> +
> + my $msg =
> + "Unit '$unit' is '$res_is_enabled' and '$res_is_active' - to prevent an excessive amount of"
> + . " audit messages being logged during the upgrade, it is recommended to stop and disable"
> + . " '$unit' beforehand."
> + . "\nThis can be achieved by running the following command:"
> + . "\n\n\tsystemctl disable --now $unit\n";
> +
> + if ($res_is_enabled eq 'enabled' || $res_is_active eq 'active') {
> + log_notice($msg);
> + return undef;
> + }
> +
> + log_pass("Unit '$unit' is '$res_is_enabled' and '$res_is_active'");
> + return undef;
> +}
> +
> sub check_misc {
> print_header("MISCELLANEOUS CHECKS");
> my $ssh_config = eval { PVE::Tools::file_get_contents('/root/.ssh/config') };
> @@ -2225,6 +2275,7 @@ sub check_misc {
> check_rrd_migration();
> check_legacy_ipam_files();
> check_legacy_sysctl_conf();
> + check_journald_audit_socket();
> }
>
> my sub colored_if {
> --
> 2.39.5
>
>
>
> _______________________________________________
> pve-devel mailing list
> pve-devel@lists.proxmox.com
> https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
>
>
>
_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [pve-devel] [PATCH pve-manager stable-8 v1 1/1] pve8to9: check if unit 'systemd-journald-audit.socket' is active
2025-08-08 8:15 ` Fabian Grünbichler
@ 2025-08-08 11:16 ` Max R. Carrara
0 siblings, 0 replies; 4+ messages in thread
From: Max R. Carrara @ 2025-08-08 11:16 UTC (permalink / raw)
To: Proxmox VE development discussion
On Fri Aug 8, 2025 at 10:15 AM CEST, Fabian Grünbichler wrote:
> On August 4, 2025 6:25 pm, Max R. Carrara wrote:
> > ... and display a `log_notice()` to the user if it is in order to
> > inform them about audit messages being logged during the upgrade,
> > as well as suggesting that they stop and disable the unit before
> > the upgrade.
> >
> > In Debian Bookworm and earlier, audit messages were not logged by
> > default [0], whereas the unit was active by default. This was changed
> > in between Bookworm and Trixie; the patch that changed the default was
> > dropped [1], whereas the unit is now disabled.
> >
> > This means that the unit will remain active during the upgrade, which
> > will in turn cause a lot of audit message log spam.
>
> so this will warn about something that happens on every system(!), but
> is only relevant for the duration of the upgrade and just makes the logs
> more noisy?
>
> I am not sure that tradeoff is worth it, it is already noted in the
> upgrade guide, so people who read that will do it proactively anyway,
> and people who notice it during the upgrade can check there as well and
> see that it's a non-issue..
Hmm yeah, I see your point...
>
> or maybe we could move it to `--full`?
No, I think it's fine to drop it then. It's in the upgrade guide and a
relatively benign thing anyway. :P
>
> adding too many warnings (and lines in general) makes it much easier to
> miss actually relevant things..
>
> in particular, after the upgrade is done it makes no sense to check this
> anymore, since the next step will be a reboot that gets rid of the
> excessive logging anyway..
^ especially because of that.
Also, I haven't seen any forum posts actually complain about the audit
log spam, so I guess it really is a non-issue.
>
> >
> > [0]: https://salsa.debian.org/systemd-team/systemd/-/commit/07daa6196f9c92be8a0f552b1416576e80d054dc
> > [1]: https://salsa.debian.org/systemd-team/systemd/-/commit/7c6ea97a1d7e438e6621c3b97ce472754fd3db43
> >
> > Signed-off-by: Max R. Carrara <m.carrara@proxmox.com>
> > ---
> > PVE/CLI/pve8to9.pm | 51 ++++++++++++++++++++++++++++++++++++++++++++++
> > 1 file changed, 51 insertions(+)
> >
> > diff --git a/PVE/CLI/pve8to9.pm b/PVE/CLI/pve8to9.pm
> > index 426fa8a6..bd300d89 100644
> > --- a/PVE/CLI/pve8to9.pm
> > +++ b/PVE/CLI/pve8to9.pm
> > @@ -2129,6 +2129,56 @@ sub check_legacy_sysctl_conf {
> > log_pass("Legacy file '$fn' exists but does not contain any settings.");
> > }
> >
> > +sub check_journald_audit_socket {
> > + my $unit = 'systemd-journald-audit.socket';
> > +
> > + log_info("Checking whether systemd unit '$unit' is active...");
> > +
> > + my ($res_is_enabled, $res_is_active) = (undef, undef);
> > +
> > + eval {
> > + run_command(
> > + ['systemctl', 'is-enabled', "$unit"],
> > + outfunc => sub {
> > + ($res_is_enabled) = @_;
> > + chomp $res_is_enabled;
> > + },
> > + noerr => 1,
> > + );
> > + };
> > +
> > + eval {
> > + run_command(
> > + ['systemctl', 'is-active', "$unit"],
> > + outfunc => sub {
> > + ($res_is_active) = @_;
> > + chomp $res_is_active;
> > + },
> > + noerr => 1,
> > + );
> > + };
> > +
> > + if (!defined($res_is_enabled) || !defined($res_is_active) || $res_is_enabled eq 'not-found') {
> > + log_skip("Unit '$unit' not found");
> > + return undef;
> > + }
> > +
> > + my $msg =
> > + "Unit '$unit' is '$res_is_enabled' and '$res_is_active' - to prevent an excessive amount of"
> > + . " audit messages being logged during the upgrade, it is recommended to stop and disable"
> > + . " '$unit' beforehand."
> > + . "\nThis can be achieved by running the following command:"
> > + . "\n\n\tsystemctl disable --now $unit\n";
> > +
> > + if ($res_is_enabled eq 'enabled' || $res_is_active eq 'active') {
> > + log_notice($msg);
> > + return undef;
> > + }
> > +
> > + log_pass("Unit '$unit' is '$res_is_enabled' and '$res_is_active'");
> > + return undef;
> > +}
> > +
> > sub check_misc {
> > print_header("MISCELLANEOUS CHECKS");
> > my $ssh_config = eval { PVE::Tools::file_get_contents('/root/.ssh/config') };
> > @@ -2225,6 +2275,7 @@ sub check_misc {
> > check_rrd_migration();
> > check_legacy_ipam_files();
> > check_legacy_sysctl_conf();
> > + check_journald_audit_socket();
> > }
> >
> > my sub colored_if {
> > --
> > 2.39.5
> >
> >
> >
> > _______________________________________________
> > pve-devel mailing list
> > pve-devel@lists.proxmox.com
> > https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
> >
> >
> >
>
>
> _______________________________________________
> pve-devel mailing list
> pve-devel@lists.proxmox.com
> https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
_______________________________________________
pve-devel mailing list
pve-devel@lists.proxmox.com
https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2025-08-08 11:15 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2025-08-04 16:25 [pve-devel] [PATCH pve-manager stable-8 v1 0/1] Note Potential Audit Log Message Flood in pve8to9 Max R. Carrara
2025-08-04 16:25 ` [pve-devel] [PATCH pve-manager stable-8 v1 1/1] pve8to9: check if unit 'systemd-journald-audit.socket' is active Max R. Carrara
2025-08-08 8:15 ` Fabian Grünbichler
2025-08-08 11:16 ` Max R. Carrara
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.