Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Bug#1033902: ulogd2 debian package missing PCAP output plugin from upstream

74 views
Skip to first unread message

Harald Welte

unread,
Apr 3, 2023, 3:10:05 PM4/3/23
to
Package: ulogd2
Version: 2.0.8-1
Severity: normal

Today - for the first time in probably 15+ years - I wanted to capture the actual packets
dropped within netfilter in a pcap file. The method I developed during my netfilter days
20 year ago for this is the PCAP output plugin of ulogd.

To my knowledge it's the only method which allows you to capture the actual binary packet
violating your iptables or nftables policy for later analysis in wireshark or other pcap
related tools.

However, I was surprised to see that the ulogd2 package both in Debian stable as well
as unstable doesn't contain the PCAP output plugin. Is that a conscious decision? I would
think it's a rather useful feature to have.

Also, the example config file contains PCAP related sections, making this even more confusing. So you
uncomment parts of the example config that gets installed (stack=log2:NFLOG,base1:BASE,pcap1:PCAP) and then it
fails due to not finding the PCAP plugin with either with

Apr 03 19:02:11 lakshmi1 ulogd[3579]: can't find requested plugin PCAP

(in the plugin auto-load case , or with

Apr 03 19:02:38 lakshmi1 ulogd[3607]: load_plugin: '/usr/lib/x86_64-linux-gnu/ulogd/ulogd_output_PCAP.so': /usr/lib/x86_64-linux-gnu/ulogd/ulogd_output_PCAP.so: cannot open shared object file: No such file or directory

(in the case one explicitly wants to load the plugin via the commented-out line from the sample config file.

Given that building the pcap plugin is enabled by default, I guess it must be explicitly disabled with
--disable-pcap in the debian package, so I guess it's a conscious decision and not an accident?

Thanks for looking into this.

-- System Information:
Debian Release: 12.0
APT prefers unstable-debug
APT policy: (500, 'unstable-debug'), (500, 'unstable')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 6.1.0-7-amd64 (SMP w/4 CPU threads; PREEMPT)
Kernel taint flags: TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=en_US.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled

Versions of packages ulogd2 depends on:
ii adduser 3.132
ii init-system-helpers 1.65.2
ii libc6 2.36-8
ii libmnl0 1.0.4-3
ii libnetfilter-acct1 1.0.3-3
ii libnetfilter-conntrack3 1.0.9-3
ii libnetfilter-log1 1.0.2-3
ii libnfnetlink0 1.0.2-2
ii lsb-base 11.6
ii sysvinit-utils [lsb-base] 3.06-3

ulogd2 recommends no packages.

Versions of packages ulogd2 suggests:
pn ulogd2-dbi <none>
pn ulogd2-json <none>
pn ulogd2-mysql <none>
pn ulogd2-pcap <none>
pn ulogd2-pgsql <none>
pn ulogd2-sqlite3 <none>

-- Configuration Files:
/etc/ulogd.conf [Errno 13] Permission denied: '/etc/ulogd.conf'

-- no debconf information

Chris Boot

unread,
Apr 4, 2023, 3:40:04 AM4/4/23
to
On 03/04/2023 19:37, Harald Welte wrote:
> However, I was surprised to see that the ulogd2 package both in Debian stable as well
> as unstable doesn't contain the PCAP output plugin. Is that a conscious decision? I would
> think it's a rather useful feature to have.

It's included in the ulogd2-pcap package, which is separate in order to
avoid the dependency on libpcap. It was this way even with ulogd 1.x.

Cheers,
Chris

--
Chris Boot
bo...@debian.org

Harald Welte

unread,
Apr 4, 2023, 4:40:04 AM4/4/23
to
On Tue, Apr 04, 2023 at 08:31:42AM +0100, Chris Boot wrote:
> On 03/04/2023 19:37, Harald Welte wrote:
> > However, I was surprised to see that the ulogd2 package both in Debian stable as well
> > as unstable doesn't contain the PCAP output plugin. Is that a conscious decision? I would
> > think it's a rather useful feature to have.
>
> It's included in the ulogd2-pcap package, which is separate in order to
> avoid the dependency on libpcap. It was this way even with ulogd 1.x.

ugh. Somehow that was too obvious. Sorry for the noise.

--
- Harald Welte <laf...@gnumonks.org> https://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
(ETSI EN 300 175-7 Ch. A6)
0 new messages