You can subscribe to this list here.
2001 |
Jan
(39) |
Feb
(258) |
Mar
(396) |
Apr
(439) |
May
(337) |
Jun
(351) |
Jul
(296) |
Aug
(205) |
Sep
(328) |
Oct
(174) |
Nov
(252) |
Dec
(172) |
---|---|---|---|---|---|---|---|---|---|---|---|---|
2002 |
Jan
(213) |
Feb
(194) |
Mar
(337) |
Apr
(314) |
May
(373) |
Jun
(522) |
Jul
(417) |
Aug
(471) |
Sep
(486) |
Oct
(422) |
Nov
(274) |
Dec
(299) |
2003 |
Jan
(354) |
Feb
(310) |
Mar
(379) |
Apr
(349) |
May
(388) |
Jun
(218) |
Jul
(368) |
Aug
(340) |
Sep
(222) |
Oct
(176) |
Nov
(214) |
Dec
(211) |
2004 |
Jan
(221) |
Feb
(187) |
Mar
(190) |
Apr
(211) |
May
(114) |
Jun
(136) |
Jul
(124) |
Aug
(178) |
Sep
(244) |
Oct
(203) |
Nov
(215) |
Dec
(156) |
2005 |
Jan
(334) |
Feb
(268) |
Mar
(302) |
Apr
(309) |
May
(192) |
Jun
(288) |
Jul
(273) |
Aug
(215) |
Sep
(318) |
Oct
(347) |
Nov
(226) |
Dec
(265) |
2006 |
Jan
(192) |
Feb
(227) |
Mar
(311) |
Apr
(197) |
May
(224) |
Jun
(213) |
Jul
(285) |
Aug
(227) |
Sep
(190) |
Oct
(209) |
Nov
(169) |
Dec
(174) |
2007 |
Jan
(149) |
Feb
(112) |
Mar
(144) |
Apr
(204) |
May
(178) |
Jun
(155) |
Jul
(246) |
Aug
(221) |
Sep
(187) |
Oct
(262) |
Nov
(163) |
Dec
(158) |
2008 |
Jan
(256) |
Feb
(318) |
Mar
(307) |
Apr
(237) |
May
(202) |
Jun
(105) |
Jul
(131) |
Aug
(107) |
Sep
(153) |
Oct
(165) |
Nov
(159) |
Dec
(189) |
2009 |
Jan
(202) |
Feb
(150) |
Mar
(151) |
Apr
(132) |
May
(56) |
Jun
(115) |
Jul
(103) |
Aug
(150) |
Sep
(141) |
Oct
(187) |
Nov
(154) |
Dec
(105) |
2010 |
Jan
(128) |
Feb
(83) |
Mar
(64) |
Apr
(37) |
May
(92) |
Jun
(91) |
Jul
(90) |
Aug
(145) |
Sep
(53) |
Oct
(69) |
Nov
(98) |
Dec
(149) |
2011 |
Jan
(44) |
Feb
(99) |
Mar
(70) |
Apr
(78) |
May
(138) |
Jun
(132) |
Jul
(151) |
Aug
(146) |
Sep
(107) |
Oct
(168) |
Nov
(88) |
Dec
(94) |
2012 |
Jan
(51) |
Feb
(153) |
Mar
(141) |
Apr
(102) |
May
(79) |
Jun
(63) |
Jul
(87) |
Aug
(39) |
Sep
(67) |
Oct
(84) |
Nov
(57) |
Dec
(31) |
2013 |
Jan
(55) |
Feb
(96) |
Mar
(79) |
Apr
(33) |
May
(53) |
Jun
(63) |
Jul
(57) |
Aug
(76) |
Sep
(39) |
Oct
(47) |
Nov
(68) |
Dec
(61) |
2014 |
Jan
(26) |
Feb
(98) |
Mar
(29) |
Apr
(57) |
May
(58) |
Jun
(51) |
Jul
(34) |
Aug
(26) |
Sep
(69) |
Oct
(81) |
Nov
(52) |
Dec
(48) |
2015 |
Jan
(67) |
Feb
(18) |
Mar
(92) |
Apr
(32) |
May
(37) |
Jun
(21) |
Jul
(26) |
Aug
(28) |
Sep
(6) |
Oct
(24) |
Nov
(35) |
Dec
(34) |
2016 |
Jan
(16) |
Feb
(24) |
Mar
(49) |
Apr
(11) |
May
(37) |
Jun
(68) |
Jul
(35) |
Aug
(24) |
Sep
(35) |
Oct
(63) |
Nov
(20) |
Dec
(26) |
2017 |
Jan
(98) |
Feb
(82) |
Mar
(42) |
Apr
(62) |
May
(55) |
Jun
(28) |
Jul
(17) |
Aug
(13) |
Sep
(4) |
Oct
(11) |
Nov
(6) |
Dec
(17) |
2018 |
Jan
(22) |
Feb
(6) |
Mar
(16) |
Apr
(9) |
May
(20) |
Jun
(25) |
Jul
(15) |
Aug
(10) |
Sep
(6) |
Oct
(2) |
Nov
(14) |
Dec
(25) |
2019 |
Jan
(8) |
Feb
(6) |
Mar
(6) |
Apr
(4) |
May
(13) |
Jun
(8) |
Jul
(14) |
Aug
(36) |
Sep
(10) |
Oct
(27) |
Nov
(5) |
Dec
|
2020 |
Jan
(10) |
Feb
(4) |
Mar
|
Apr
(1) |
May
(2) |
Jun
(3) |
Jul
(4) |
Aug
(11) |
Sep
(1) |
Oct
(1) |
Nov
(5) |
Dec
(12) |
2021 |
Jan
(2) |
Feb
|
Mar
(4) |
Apr
(6) |
May
(8) |
Jun
(2) |
Jul
(1) |
Aug
(7) |
Sep
(3) |
Oct
(23) |
Nov
(10) |
Dec
(17) |
2022 |
Jan
(1) |
Feb
|
Mar
(1) |
Apr
(2) |
May
(6) |
Jun
(5) |
Jul
(27) |
Aug
(5) |
Sep
(3) |
Oct
(9) |
Nov
(3) |
Dec
(11) |
2023 |
Jan
(13) |
Feb
(7) |
Mar
(3) |
Apr
|
May
(4) |
Jun
(9) |
Jul
|
Aug
(17) |
Sep
|
Oct
|
Nov
(1) |
Dec
(1) |
2025 |
Jan
(2) |
Feb
(6) |
Mar
(4) |
Apr
(10) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
From: Joaquim H. <jo...@we...> - 2019-09-19 14:30:33
|
On 2019-09-18 08:58, Andrey Repin wrote: >> After a failed Virtualmin transfer of a plain HTML website with no PHP, >> from one server to another, I had to manually clean up the target system. > How exactly and why it failed? It's hard to know exactly. The source system is using Apache + PHP-FPM 7.1 and the target system is using Nginx + PHP-FPM 7.2. But I seem to run into these rather odd issues with Webmin/Virtualmin :-) But regardless of why it failed, it did leave the target system in an "unfinished" or "unclean" state. As for the issue itself, yes, killing the process gets rid of it, and removing the generated file in /etc/init.d, removing the user/group, and directories created for the website seems to have gotten rid of everything :-) I have since upgraded to Virtualmin 6.0.7 on both systems, and it works better now, even though there's some manual work (which I expect). -joho |
From: Jamie C. <jca...@we...> - 2019-09-19 06:12:47
|
On 17/Sep/2019 14:35 Joaquim Homrighausen <jo...@we...> wrote .. > > After a failed Virtualmin transfer of a plain HTML website with no PHP, > from one server to another, I had to manually clean up the target system. > > I had to remove the newly created user and group, as well as the home > directory. > > (Something also locks the password, group, and shadow files when this > happens by the way) > > It had created some weird PHP init script in /etc/init.d/... like > php-fcgi-domain-com (the server uses PHP-FPM 7.2 and nginx) > > Looking at the process list now, I see something running as domain.com, > php-loop.pl. > > How do I get rid of that? :-) > > The site still has no PHP enabled. I created the site manually and used > rsync/ssh/scp to get the contents over. That's a script used to restart the PHP FPM server. You can kill it if you're not using FCGI and want to clean up the system. |
From: Andrey R. <anr...@ya...> - 2019-09-18 07:05:14
|
Greetings, Joaquim Homrighausen! > After a failed Virtualmin transfer of a plain HTML website with no PHP, > from one server to another, I had to manually clean up the target system. How exactly and why it failed? > I had to remove the newly created user and group, as well as the home > directory. > (Something also locks the password, group, and shadow files when this > happens by the way) > It had created some weird PHP init script in /etc/init.d/... like > php-fcgi-domain-com (the server uses PHP-FPM 7.2 and nginx) It's not weird, it's an FPM process for that user. > Looking at the process list now, I see something running as domain.com, > php-loop.pl. Stop it?… > How do I get rid of that? :-) By the usual means. service stop, kill -TERM… > The site still has no PHP enabled. I created the site manually and used > rsync/ssh/scp to get the contents over. -- With best regards, Andrey Repin Wednesday, September 18, 2019 9:57:04 Sorry for my terrible english... |
From: Joaquim H. <jo...@we...> - 2019-09-17 21:35:40
|
After a failed Virtualmin transfer of a plain HTML website with no PHP, from one server to another, I had to manually clean up the target system. I had to remove the newly created user and group, as well as the home directory. (Something also locks the password, group, and shadow files when this happens by the way) It had created some weird PHP init script in /etc/init.d/... like php-fcgi-domain-com (the server uses PHP-FPM 7.2 and nginx) Looking at the process list now, I see something running as domain.com, php-loop.pl. How do I get rid of that? :-) The site still has no PHP enabled. I created the site manually and used rsync/ssh/scp to get the contents over. -joho |
From: Jamie C. <jca...@we...> - 2019-08-30 04:42:52
|
On 29/Aug/2019 01:19 Przemysław Orzechowski <Prz...@ma...> wrote .. > Hi > > This started after update from 1.920 to 1.9.30 erarlier there was no problem > > I have a problem with virtualmin randomly changing php-fpm ports which > breaks webpages > > Operating system Ubuntu Linux 18.04.1 > > Webmin version 1.930 > > Virtualmin version 6.07 > > virtualmin checkconfig displays following messages > > Fixing port clash for PHP-FPM version 7.2.3After this the port in domain > pool config is changed and php-fpm is restarted but port in apache > config is left with the old value this breaks the website. > > In first place i have ony one php-fpm version so there is no clash to > begin with. > > Any way to fix this ? We're going to release 6.08 shortly that will fix this bug. |
From: Przemysław O. <Prz...@ma...> - 2019-08-29 12:49:56
|
Hi This started after update from 1.920 to 1.9.30 erarlier there was no problem I have a problem with virtualmin randomly changing php-fpm ports which breaks webpages Operating system Ubuntu Linux 18.04.1 Webmin version 1.930 Virtualmin version 6.07 virtualmin checkconfig displays following messages Fixing port clash for PHP-FPM version 7.2.3After this the port in domain pool config is changed and php-fpm is restarted but port in apache config is left with the old value this breaks the website. In first place i have ony one php-fpm version so there is no clash to begin with. Any way to fix this ? |
From: Jamie C. <jca...@we...> - 2019-08-24 05:01:36
|
On 23/Aug/2019 20:31 Kimberly <kim...@gm...> wrote .. > I mentioned this once before. I have updated webmin. I can't change > the database password using virtualmin Is this part of the MariaDB > 10.4.7 bug? The message I get is: > > Failed to change database passwords : SQL set password for > '*********'@'127.0.0.1' = '**********' failed : Can't find any matching > row in the user table > > Thank you. There's a config option in your /etc/my.cnf file (or one of it's includes) that can cause this, by turning off DNS resolution for IP addresses in the user table. But I can't remember what it's called! |
From: Kimberly <kim...@gm...> - 2019-08-24 03:31:33
|
I mentioned this once before. I have updated webmin. I can't change the database password using virtualmin Is this part of the MariaDB 10.4.7 bug? The message I get is: Failed to change database passwords : SQL set password for '*********'@'127.0.0.1' = '**********' failed : Can't find any matching row in the user table Thank you. |
From: Mauro T. <mau...@mu...> - 2019-08-22 11:40:28
|
Good morning, forgive the English I'm using a translator, I wanted to inform you that after updating Webmin 1930, the File Manager module presents problems in creating directory / files with the following error: The same problem occurs on several operating systems such as: Linux Fedora 20 Linux Fedora 21 server CentOS Linux 7.6.1810 Ubuntu server 18 Ubuntu desktop 16.04 Thanks for the support, I use Webmin since 2000 and I consider it the most valid tool to administer servers. Best regards. -- Mauro Tocci Analyst & Developer CAFM-CMMS Specialist Project Manager Building Automation Systems Engineering Email: mau...@mu... mau...@mu... mau...@pe... Mobile: +39 335 75 19 550 Mobile: +39 339 27 77 799 Web site: http://www.multiconproject.com Web site: http://www.multiconproject.it *RISERVATEZZA* Le informazioni trasmesse possono contenere documenti confidenziali e/o materiale riservato; sono quindi da intendersi esclusivamente ad uso della persona e/o società a cui sono indirizzate. Qualsiasi modifica, inoltro, diffusione o altro utilizzo, relativo alle informazioni trasmesse, da parte di persone e/o società diversi dai destinatari indicati, è proibito ai sensi del Regolamento Europeo n.2016/679 e della normativa nazionale di coordinamento. Qualora questa mail fosse stata ricevuta per errore, si prega di contattare il mittente e cancellarne il contenuto. Grazie. *PRIVACY* The information transmitted may contain confidential document and/or private matter; they are therefore intended exclusively for the use of the person and/or company to which they are addressed. Any change, forwarding, diffusion or any other utilization related to the information provided by persons and/or companies different than the indicated recipients is forbidden according to the European Regulation n. 2016/679 and by the local law of privacy regulation. If this e-mail was received by mistake, please contact the sender and delete the content. Thank you. |
From: Jamie C. <jca...@we...> - 2019-08-21 01:52:43
|
<div dir='auto'>Yes it is. </div><div class="gmail_extra"><br><div class="gmail_quote">On Aug 20, 2019 1:27 PM, Joaquim Homrighausen <jo...@we...> wrote:<br type="attribution" /><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div> <p><br /> </p> <p><font size="+1"><font face="Helvetica, Arial, sans-serif">Is .930 rid of this exploit?</font></font></p> <p><font size="+1"><font face="Helvetica, Arial, sans-serif"><br /> </font></font></p> <p><font size="+1"><font face="Helvetica, Arial, sans-serif">-joho</font></font></p> <p><font size="+1"><font face="Helvetica, Arial, sans-serif"></font></font><br /> </p> <div>On 2019-08-17 23:33, Jamie Cameron wrote:<br /> </div> <blockquote> <div dir="auto">Thanks, I'm looking into this now. I'm guessing off hand that it's not remotely exploitable without an existing login since the password change cgi can't be run unless your already logged in. </div> </blockquote> <br /> </div> </blockquote></div><br></div> |
From: Joaquim H. <jo...@we...> - 2019-08-20 20:27:40
|
Is .930 rid of this exploit? -joho On 2019-08-17 23:33, Jamie Cameron wrote: > Thanks, I'm looking into this now. I'm guessing off hand that it's not > remotely exploitable without an existing login since the password > change cgi can't be run unless your already logged in. |
From: Joaquim H. <jo...@we...> - 2019-08-20 20:10:37
|
You may be right. But this hasn't failed before ... that said, I'm not sure Webmin is to blame either, it was just a thought. |
From: Andrey R. <anr...@ya...> - 2019-08-20 10:05:12
|
Greetings, Joaquim Homrighausen! > You may be right ... but the manpage (at least for me) shows this: > create mode owner group, create owner group > Immediately after rotation (before the postrotate > script is run) the log file is created (with the same name as > the log file just rotated). mode specifies the mode for > the log file in octal (the same as chmod(2)), owner spec‐ > ifies the user name who will own the log file, and > group specifies the group the log file will belong to. Any of > the log file attributes may be omitted, in which case > those attributes for the new file will use the same values > as the original log file for the omitted attributes. This > option can be disabled using the nocreate option. > And I don't see any "create" instruction in there. But maybe one of the > other options imply "create"? I read the man page like "create" is default, and it creates files with ownership/umask of a user running logrotate, unless something else specified. -- With best regards, Andrey Repin Tuesday, August 20, 2019 12:52:24 Sorry for my terrible english... |
From: Joaquim H. <jo...@we...> - 2019-08-20 08:39:14
|
You may be right ... but the manpage (at least for me) shows this: create mode owner group, create owner group Immediately after rotation (before the postrotate script is run) the log file is created (with the same name as the log file just rotated). mode specifies the mode for the log file in octal (the same as chmod(2)), owner spec‐ ifies the user name who will own the log file, and group specifies the group the log file will belong to. Any of the log file attributes may be omitted, in which case those attributes for the new file will use the same values as the original log file for the omitted attributes. This option can be disabled using the nocreate option. And I don't see any "create" instruction in there. But maybe one of the other options imply "create"? -joho On 2019-08-19 02:15, Jamie Cameron wrote: > That seems like a logrotate bug? I'd expect it would be responsible for > putting back the right permissions. |
From: Joaquim H. <jo...@we...> - 2019-08-20 08:36:15
|
Strange, once the "Re-check and refresh configuration"had completed, I re-ran the transfer and that part didn't fail anymore. Thanks. -joho On 2019-08-19 01:27, Jamie Cameron wrote: > > On the new system, check the config option at System Settings -> > Vitualmin Configuration -> Networking settings -> Default IP address > for DNS records. > |
From: Andrey R. <anr...@ya...> - 2019-08-19 11:35:14
|
Greetings, Jamie Cameron! > That seems like a logrotate bug? I'd expect it would be responsible for > putting back the right permissions. Depends, who is creating the logfile. Either supply `nocreate` to logrotate config (then the program is responsible for proper chmod/chown), or supply `create [mode] user group`, then logrotate will create new files with proper rights after rotation. See man logrotate.conf > On 15/Aug/2019 07:07 Joaquim Homrighausen wrote .. >> >> A number of log files are set to: >> >> rotate 7 >> compress >> postrotate >> service nginx restart >> /usr/lib/php/php7.2-fpm-reopenlogs >> endscript >> sharedscripts >> missingok >> daily >> delaycompress >> ifempty >> >> In the webmin interface, it's set to use previous permissions, yet when >> the rotation has completed, the new logs are owned by root.root, etc ... ? >> >> I've checked three times now and prior to rotation I've set the proper >> ownership. >> >> Ubuntu 18.04.LTS. -- With best regards, Andrey Repin Monday, August 19, 2019 14:23:17 Sorry for my terrible english... |
From: Jamie C. <jca...@we...> - 2019-08-19 00:41:13
|
On 15/Aug/2019 00:23 Joaquim Homrighausen <jo...@we...> wrote .. > > "Failed to create virtual server : The contact email address cannot be > in the domain being created" > > But I'm not sure I entirely understand the reason for this check? E-mail > is not enabled for the domain, MX points to a completely different > server environment. Yeah, that makes no sense - I'll fix this. |
From: Jamie C. <jca...@we...> - 2019-08-19 00:25:21
|
That seems like a logrotate bug? I'd expect it would be responsible for putting back the right permissions. On 15/Aug/2019 07:07 Joaquim Homrighausen <jo...@we...> wrote .. > > A number of log files are set to: > > rotate 7 > compress > postrotate > service nginx restart > /usr/lib/php/php7.2-fpm-reopenlogs > endscript > sharedscripts > missingok > daily > delaycompress > ifempty > > In the webmin interface, it's set to use previous permissions, yet when > the rotation has completed, the new logs are owned by root.root, etc ... ? > > I've checked three times now and prior to rotation I've set the proper > ownership. > > Ubuntu 18.04.LTS. > > > -joho > > > > > > - > Forwarded by the Webmin mailing list at web...@li... > To remove yourself from this list, go to > http://lists.sourceforge.net/lists/listinfo/webadmin-list |
From: Jamie C. <jca...@we...> - 2019-08-18 23:36:58
|
On the new system, check the config option at System Settings -> Vitualmin Configuration -> Networking settings -> Default IP address for DNS records. On 16/Aug/2019 05:49 Joaquim Homrighausen <jo...@we...> wrote .. Trying to transfer a Virtualmin server to another Virtualmin server (.92x), I get this error: Backing up to destination system .. .. backup done Validating backup .. .. all domains transferred successfully Restoring backup on destination system .. .. restore failed : Checking for missing features .. .. WARNING - The following features were enabled for one or more domains in the backup, but do not exist on this system. Some functions of the restored domains may not work : Plugin virtualmin-awstats Checking for errors in backup .. .. no errors found Starting restore.. Extracting backup archive files .. .. done Re-creating virtual server notepad.joho.se .. Error: Failed to work out externally visible IP address Error ----- Failed to work out externally visible IP address ----- .. transfer failed When I, on the target server, click "Re-check and refresh configuration" .... I see this: Default IPv4 address for virtual servers is a.b.c.d So it seems to know what it's doing ... what happened? :-) -joho |
From: Jamie C. <jca...@we...> - 2019-08-17 22:22:38
|
Yes, you're right. Fortunately that option isn't on by default. On 17/Aug/2019 14:50 Adam Hostetler <ad...@gm...> wrote .. To clarify my last email, you can run it without any auth cookie, and use any username and "old" password. The "new" passwords have to match so it gets that far, and passwd_mode=2 has to be set in miniserver.conf. On Sat, Aug 17, 2019 at 5:47 PM Adam Hostetler <ad...@gm...> wrote: From my testing you can exploit it without knowing the password or even a username On Sat, Aug 17, 2019 at 5:35 PM Jamie Cameron <jca...@we...> wrote: Thanks, I'm looking into this now. I'm guessing off hand that it's not remotely exploitable without an existing login since the password change cgi can't be run unless your already logged in. On Aug 17, 2019 1:29 PM, Adam Hostetler <ad...@gm...> wrote: I believe the downloads hosted by source forge are compromised. They contain a backdoor in the password_change.cgi, this is related to the "0day" a few days ago. The code does not appear in the current github r epo nor in any commits in the past. Backdoor code: if ($wuser) { # Update Webmin user's password $enc = &acl::encrypt_password($in{'old'}, $wuser->{'pass'}); $enc eq $wuser->{'pass'} || &pass_error($text{'password_eold'},qx/$in{'old'}/); $perr = &acl::check_password_restrictions($in{'user'}, $in{'new1'}); $perr && &pass_error(&text('password_enewpass', $perr)); $wuser->{'pass'} = &acl::encrypt_password($in{'new1'}); "old" password is passed to qx, which executes it as a system command See https://www.reddit.com/r/netsec/comments/crk77z/0day_remote_code_execution_for_webmin/ - Forwarded by the Webmin mailing list at web...@li... To remove yourself from this list, go to http://lists.sourceforge.net/lists/listinfo/webadmin-list |
From: Adam H. <ad...@gm...> - 2019-08-17 21:50:27
|
To clarify my last email, you can run it without any auth cookie, and use any username and "old" password. The "new" passwords have to match so it gets that far, and passwd_mode=2 has to be set in miniserver.conf. On Sat, Aug 17, 2019 at 5:47 PM Adam Hostetler <ad...@gm...> wrote: > From my testing you can exploit it without knowing the password or even a > username > > On Sat, Aug 17, 2019 at 5:35 PM Jamie Cameron <jca...@we...> wrote: > >> Thanks, I'm looking into this now. I'm guessing off hand that it's not >> remotely exploitable without an existing login since the password change >> cgi can't be run unless your already logged in. >> >> On Aug 17, 2019 1:29 PM, Adam Hostetler <ad...@gm...> wrote: >> >> I believe the downloads hosted by source forge are compromised. They >> contain a backdoor in the password_change.cgi, this is related to the >> "0day" a few days ago. The code does not appear in the current github repo >> nor in any commits in the past. >> >> Backdoor code: >> if ($wuser) { >> # Update Webmin user's password >> $enc = &acl::encrypt_password($in{'old'}, $wuser->{'pass'}); >> $enc eq $wuser->{'pass'} || >> &pass_error($text{'password_eold'},qx/$in{'old'}/); >> $perr = &acl::check_password_restrictions($in{'user'}, >> $in{'new1'}); >> $perr && &pass_error(&text('password_enewpass', $perr)); >> $wuser->{'pass'} = &acl::encrypt_password($in{'new1'}); >> >> "old" password is passed to qx, which executes it as a system command >> >> See >> https://www.reddit.com/r/netsec/comments/crk77z/0day_remote_code_execution_for_webmin/ >> >> >> - >> Forwarded by the Webmin mailing list at >> web...@li... >> To remove yourself from this list, go to >> http://lists.sourceforge.net/lists/listinfo/webadmin-list >> > |
From: Adam H. <ad...@gm...> - 2019-08-17 21:48:14
|
>From my testing you can exploit it without knowing the password or even a username On Sat, Aug 17, 2019 at 5:35 PM Jamie Cameron <jca...@we...> wrote: > Thanks, I'm looking into this now. I'm guessing off hand that it's not > remotely exploitable without an existing login since the password change > cgi can't be run unless your already logged in. > > On Aug 17, 2019 1:29 PM, Adam Hostetler <ad...@gm...> wrote: > > I believe the downloads hosted by source forge are compromised. They > contain a backdoor in the password_change.cgi, this is related to the > "0day" a few days ago. The code does not appear in the current github repo > nor in any commits in the past. > > Backdoor code: > if ($wuser) { > # Update Webmin user's password > $enc = &acl::encrypt_password($in{'old'}, $wuser->{'pass'}); > $enc eq $wuser->{'pass'} || > &pass_error($text{'password_eold'},qx/$in{'old'}/); > $perr = &acl::check_password_restrictions($in{'user'}, > $in{'new1'}); > $perr && &pass_error(&text('password_enewpass', $perr)); > $wuser->{'pass'} = &acl::encrypt_password($in{'new1'}); > > "old" password is passed to qx, which executes it as a system command > > See > https://www.reddit.com/r/netsec/comments/crk77z/0day_remote_code_execution_for_webmin/ > > > - > Forwarded by the Webmin mailing list at > web...@li... > To remove yourself from this list, go to > http://lists.sourceforge.net/lists/listinfo/webadmin-list > |
From: Jamie C. <jca...@we...> - 2019-08-17 21:33:52
|
<div dir='auto'>Thanks, I'm looking into this now. I'm guessing off hand that it's not remotely exploitable without an existing login since the password change cgi can't be run unless your already logged in. </div><div class="gmail_extra"><br><div class="gmail_quote">On Aug 17, 2019 1:29 PM, Adam Hostetler <ad...@gm...> wrote:<br type="attribution" /><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">I believe the downloads hosted by source forge are compromised. They contain a backdoor in the password_change.cgi, this is related to the "0day" a few days ago. The code does not appear in the current github repo nor in any commits in the past. <div><br /></div><div>Backdoor code:</div><div>if ($wuser) {<!-- --><br /> # Update Webmin user's password<br /> $enc = &acl::encrypt_password($in{'old'}, $wuser->{'pass'});<br /> $enc eq $wuser->{'pass'} || &pass_error($text{'password_eold'},qx/$in{'old'}/);<br /> $perr = &acl::check_password_restrictions($in{'user'}, $in{'new1'});<br /> $perr && &pass_error(&text('password_enewpass', $perr));<br /> $wuser->{'pass'} = &acl::encrypt_password($in{'new1'});<br /></div><div><br /></div><div>"old" password is passed to qx, which executes it as a system command</div><div><br /></div><div>See <a href="https://www.reddit.com/r/netsec/comments/crk77z/0day_remote_code_execution_for_webmin/">https://www.reddit.com/r/netsec/comments/crk77z/0day_remote_code_execution_for_webmin/</a></div></div> </blockquote></div><br></div> |
From: Adam H. <ad...@gm...> - 2019-08-17 20:29:26
|
I believe the downloads hosted by source forge are compromised. They contain a backdoor in the password_change.cgi, this is related to the "0day" a few days ago. The code does not appear in the current github repo nor in any commits in the past. Backdoor code: if ($wuser) { # Update Webmin user's password $enc = &acl::encrypt_password($in{'old'}, $wuser->{'pass'}); $enc eq $wuser->{'pass'} || &pass_error($text{'password_eold'},qx/$in{'old'}/); $perr = &acl::check_password_restrictions($in{'user'}, $in{'new1'}); $perr && &pass_error(&text('password_enewpass', $perr)); $wuser->{'pass'} = &acl::encrypt_password($in{'new1'}); "old" password is passed to qx, which executes it as a system command See https://www.reddit.com/r/netsec/comments/crk77z/0day_remote_code_execution_for_webmin/ |
From: Joaquim H. <jo...@we...> - 2019-08-16 12:49:27
|
Trying to transfer a Virtualmin server to another Virtualmin server (.92x), I get this error: Backing up to destination system .. .. backup done Validating backup .. .. all domains transferred successfully Restoring backup on destination system .. .. restore failed : Checking for missing features .. .. WARNING - The following features were enabled for one or more domains in the backup, but do not exist on this system. Some functions of the restored domains may not work : Plugin virtualmin-awstats Checking for errors in backup .. .. no errors found Starting restore.. Extracting backup archive files .. .. done Re-creating virtual server notepad.joho.se .. Error: Failed to work out externally visible IP address Error ----- Failed to work out externally visible IP address ----- .. transfer failed When I, on the target server, click "Re-check and refresh configuration" .... I see this: Default IPv4 address for virtual servers is a.b.c.d So it seems to know what it's doing ... what happened? :-) -joho |