Your message dated Thu, 20 Apr 2017 16:40:29 +0200 with message-id <cafx5sbwtzyt8edsbge_6_arrbexehv5-qxwrwrq2ho-ybxk...@mail.gmail.com> and subject line Re: [Pkg-samba-maint] Bug#860779: smbclient: installation of smbclient appears to install and run samba server has caused the Debian Bug report #860779, regarding smbclient: installation of smbclient appears to install and run samba server to be marked as done.
This means that you claim that the problem has been dealt with. If this is not the case it is now your responsibility to reopen the Bug report if necessary, and/or fix the problem forthwith. (NB: If you are a system administrator and have no idea what this message is talking about, this may indicate a serious mail system misconfiguration somewhere. Please contact [email protected] immediately.) -- 860779: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=860779 Debian Bug Tracking System Contact [email protected] with problems
--- Begin Message ---Package: smbclient Version: 2:4.2.14+dfsg-0+deb8u5 Severity: normal I do not think that installing client software should, without much notice, install and activate the associated server. But that seems to be what happens with smbclient. Among other things, this seems an unnecessary security risk. * What led up to the situation? I wanted to access a samba share being served by another machine. So I installed smbclient using aptitude, accepting the defaults, which I noticed included samba (the main package) and a lot of other things. * What exactly did you do (or not do) that was effective (or ineffective)? I first ran systemtcl status samba which seemed to indicate the server was not running. However, the samba logs and netstat indicated it was, as did /etc/init.d/samba status * What was the outcome of this action? Big picture: installing the client resulted in a running server on my machine. As for my diagnostics: ross@ross-node1:/tmp$ systemctl status samba ● samba.service Loaded: masked (/dev/null) Active: inactive (dead) ross@ross-node1:/tmp$ /etc/init.d/samba status ● nmbd.service - LSB: start Samba NetBIOS nameserver (nmbd) Loaded: loaded (/etc/init.d/nmbd) Active: active (running) since Wed 2017-04-19 16:36:11 PDT; 41min ago CGroup: /system.slice/nmbd.service └─52137 /usr/sbin/nmbd -D ● smbd.service - LSB: start Samba SMB/CIFS daemon (smbd) Loaded: loaded (/etc/init.d/smbd) Active: active (running) since Wed 2017-04-19 16:36:10 PDT; 41min ago CGroup: /system.slice/smbd.service ├─52068 /usr/sbin/smbd -D └─52072 /usr/sbin/smbd -D * What outcome did you expect instead? Well, I wasn't too surprised to find the server running given the packages installed. But my original expectation was that I could install the client without getting a server started. Our network admin is pretty strict and takes a dim view of random services running on the network. So I'm going to remove all the packages for now. As a side note, I find the current interaction of samba and systemd to be mysterious and undocumented. I did find bug 740942 which shed some light (namely that the samba.service link to /dev/null sort of tells systemd to ignore the package), but I remain puzzled. The end of that bug links to http://git.debian.org/?p=pkg-samba/samba.git;a=commitdiff;h=8828d90 but that link doesn't seem to work anymore. Then again, I find systemd to be generally mysterious. In particular, I don't know what the proper way to disable the services is. I do notice that smbclient relies on configuration parameters in smb.conf, and so it may be that getting a "pure" client is technically difficult. -- System Information: Debian Release: 8.7 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Kernel: Linux 3.16.0-4-amd64 (SMP w/40 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages smbclient depends on: ii dpkg 1.17.27 ii libarchive13 3.1.2-11+deb8u3 ii libbsd0 0.7.0-2 ii libc6 2.19-18+deb8u7 ii libpopt0 1.16-10 ii libreadline6 6.3-8+b3 ii libsmbclient 2:4.2.14+dfsg-0+deb8u5 ii libtalloc2 2.1.2-0+deb8u1 ii libtevent0 0.9.28-0+deb8u1 ii samba-common 2:4.2.14+dfsg-0+deb8u5 ii samba-libs 2:4.2.14+dfsg-0+deb8u5 smbclient recommends no packages. Versions of packages smbclient suggests: ii cifs-utils 2:6.4-1 pn heimdal-clients <none> -- no debconf information
--- End Message ---
--- Begin Message ---2017-04-20 2:31 GMT+02:00 Ross Boylan <[email protected]>: > Package: smbclient > Version: 2:4.2.14+dfsg-0+deb8u5 > Severity: normal > > I do not think that installing client software should, without much > notice, install and activate the associated server. But that seems to > be what happens with smbclient. Among other things, this seems an > unnecessary security risk. I can't reproduce this problem. Thus I'm closing this bug. Try "aptitude why samba". But, we have the problem of winbind depending on samba (#732604). This is fixed in stretch and is a major change that can't happen in jessie. > > * What led up to the situation? > I wanted to access a samba share being served by another machine. > So I installed smbclient using aptitude, accepting the defaults, > which I noticed included samba (the main package) and a lot of > other things. Alternatively, you can use cifs-utils and "mount -t cifs". > > * What exactly did you do (or not do) that was effective (or > ineffective)? > I first ran systemtcl status samba which seemed to indicate the > server was not running. However, the samba logs and netstat > indicated it was, as did /etc/init.d/samba status samba.service is not usefull. it was kept, but this was a bad idea (we should have done like the gitlab.target and gitlab.service, which behave correctly as a meta service) > > * What was the outcome of this action? > Big picture: installing the client resulted in a running server on > my machine. As for my diagnostics: > ross@ross-node1:/tmp$ systemctl status samba > ● samba.service > Loaded: masked (/dev/null) > Active: inactive (dead) > ross@ross-node1:/tmp$ /etc/init.d/samba status > ● nmbd.service - LSB: start Samba NetBIOS nameserver (nmbd) > Loaded: loaded (/etc/init.d/nmbd) > Active: active (running) since Wed 2017-04-19 16:36:11 PDT; 41min ago > CGroup: /system.slice/nmbd.service > └─52137 /usr/sbin/nmbd -D > ● smbd.service - LSB: start Samba SMB/CIFS daemon (smbd) > Loaded: loaded (/etc/init.d/smbd) > Active: active (running) since Wed 2017-04-19 16:36:10 PDT; 41min ago > CGroup: /system.slice/smbd.service > ├─52068 /usr/sbin/smbd -D > └─52072 /usr/sbin/smbd -D You'd better use "systemctl status smbd nmbd". > * What outcome did you expect instead? > Well, I wasn't too surprised to find the server running given > the packages installed. But my original expectation was that I > could install the client without getting a server started. > > Our network admin is pretty strict and takes a dim view of random > services running on the network. So I'm going to remove all the > packages for now. > > As a side note, I find the current interaction of samba and systemd to > be mysterious and undocumented. I did find bug 740942 which shed some > light (namely that the samba.service link to /dev/null sort of tells > systemd to ignore the package), but I remain puzzled. The end of that > bug links to > http://git.debian.org/?p=pkg-samba/samba.git;a=commitdiff;h=8828d90 > but that link doesn't seem to work anymore. https://anonscm.debian.org/cgit/pkg-samba/samba.git/commit/?id=8828d90 > Then again, I find > systemd to be generally mysterious. It is more a misunderstanding of systemd by us, packagers of samba (or more precisely myself). As systemd isolate a service in a cgroup we had to move nmbd and smbd in different cgroups. This led to #740942 that I fixed by masking the samba service. This is still not fixed in stretch. > In particular, I don't know what the proper way to disable the > services is. systemctl disable smbd nmbd # or mask, see http://0pointer.de/blog/projects/three-levels-of-off.html > I do notice that smbclient relies on configuration parameters in > smb.conf, and so it may be that getting a "pure" client is technically > difficult. Regards -- Mathieu Parent
--- End Message ---
