Discuss November 2014

discuss@lists.centos.org

182 participants
199 discussions

cr repo firefox el6.6 crashes on 6.5
by Robert Arkiletian 18 Nov '14

18 Nov '14

Noticed firefox is one of the only critical sec. updates from 6.6. So I manually downloaded the rpm from cr repo and rpm -Uvh firefox-31.2.0-3.el6.centos.x86_64.rpm but it crashes with /usr/lib64/firefox/firefox: symbol lookup error: /usr/lib64/firefox/libxul.so: undefined symbol: gdk_window_get_visual

3 3

FreeIPA 4.1.1 does not install on Cntos7
by Cosme Faria Corrêa 18 Nov '14

18 Nov '14

Hi, We're looking to run freeipa 4.1.1 on CentOS 7. 1. after include: * mkosek-freeipa-epel-7.repo 2. Write this: * yum install freeipa-server 3. I get this error: * Error: Package: pki-base-10.2.0-3.el7.centos.noarch (mkosek-freeipa) * Requires: jackson-jaxrs-json-provider I would appreciate any ideas. -- Cosme Faria Corrêa

2 1

automated stopserver & shutdown?
by Hadi Motamedi 18 Nov '14

18 Nov '14

Dear All In an environment , I have 20 centos servers running together . For shutting them down , I need to issue the followings on each of the servers : #./stopServer #init 0 This is cumbersome to try to issue these on huge amount of servers to keep them safely going shutdown. Can you please let me know how can I automate it and say let just one server send stop processes and shutdown commands to the other ones and then goes shutdown himself? Thank you

4 7

Anyone have a Brother multifunction working on Centos 7?
by Negative 18 Nov '14

18 Nov '14

I have a Brother MFC 7360N, and it is refusing to print. Given a job, the printer state shows, "Sending data to printer"; the file is in the spool dir; the job queue is empty, and nothing more happens. I had used the Brother driver installer, and ran into a 32-bit problem. The rpm runs braddprinter, which fails looking for /lib/ld-linux.so.2. I don't know what braddprinter does, but I then tried setting up the printer through the system printer config. Everything seems to go ok. A driver is in /usr/share/cups. I tried to install glibc.i686, but it wants glibc-common. I have the 64-bit glibc-common package but do not see a i686 one available. Thanks in adv.

4 4

Installing old version of firefox
by Larry Martell 18 Nov '14

18 Nov '14

I need to install firefox version 24.3.0 on centos version 6.5. Anyone know how I can do that? I tried downloading that version directly and it fails with: XPCOMGlueLoad error for file /usr/local/lib64/firefox/libxul.so: libXrender.so.1: cannot open shared object file: No such file or directory Couldn't load XPCOM.

8 21

x2go - F20 to Centos 7
by David A. De Graaf 17 Nov '14

17 Nov '14

All my attempts to access a Centos 7 machine from Fedora 20 using x2go have failed with these error messages in /var/log/secure: Nov 17 16:42:08 datium sshd[4175]: Accepted publickey for dad from 127.0.0.1 port 43005 ssh2: RSA c5:d0:02:e7:90:ce:8b:bb:65:b2:78:66:29:8c:c8:15 Nov 17 16:42:08 datium sshd[4175]: pam_unix(sshd:session): session opened for user dad by (uid=0) Nov 17 16:42:08 datium sshd[4185]: subsystem request for sftp by user dad Nov 17 16:42:09 datium sshd[4185]: Received disconnect from 127.0.0.1: 11: disconnected by user Nov 17 16:42:09 datium sshd[4175]: pam_unix(sshd:session): session closed for user dad The Centos 7 machine is virtualized in kvm, running on another F20 machine, but that shouldn't matter. ssh is configured to allow access via public/private keys and works fine. sshd seems to have a problem with "subsystem request for sftp", whatever that means, but I have verified that I can run 'sftp datwizkvm' and get access to my home directory there. When x2go has negotiated the public key transaction, an XFCE window pops up for about 1 sec and then vanishes. This leads me to speculate that the problem is more to do with starting X than making the connection. In the opposite direction, x2goclient on the Centos 7 machine will happily open and run an XFCE display of the Fedora 20 machine. Does anyone have insight about why the connection is asymmetric? Has anyone experienced this kind of failure? Any hints how to fix it? -- David A. De Graaf DATIX, Inc. Hendersonville, NC dad(a)datix.us www.datix.us "Wagner's music is better than it sounds." -Mark Twain

1 0

Re: [CentOS] Remote Desktop Recommendations (2)
by david 17 Nov '14

17 Nov '14

At 09:30 AM 11/17/2014, you wrote: >On Mon, Nov 17, 2014 at 11:06 AM, david <david(a)daku.org> wrote: > > > > > Thanks, Les, for that detail. I knew that x2go had a known bug, but didn't > > know the name of the missing package. I explicitly installed it, but I > > still get failure asking for the MATE desktop. To summarize, my client > > shows the X2go screen, then it goes away. > >The version that is working for me is >x2goserver-4.0.1.18-3.el7.x86_64. If that is not what you have, try: >yum--enablerepo=epel-testing update x2goserver >You might also have to do something to clean up the old sessions that >the broken version had started - I can't remember what else I might >have done. Also, if you can, make sure that MATE is working at the >console, and for the first attempts turn off the drive mapping and >sound to have fewer things that can break. > > > In response to others that suggested VNC, I used to use it, but was > > disappointed with the speed when used remotely. On an internal > LAN connect, > > it worked fine, but otherwise it became painful to use. NX became my next > > choice for Centos 6, but the x2go package became my favorite because of the > > ease of installation -- (nx needed lots of other things at setup, and > > appears unavailable for C7). > > > >So far I am still mostly using NX/freenx on a C6 system as my main >desktop with occasional 'ssh -Y ...' connections from there to C7 for >GUI programs. I think x2go will be a usable replacement, but the >windows client has some conflict with 'synergy' which I also use to >share the keyboard/mouse with an adjacent laptop so I haven't switched >yet. Starting from a text login, what's the "startx" equivalent to start MATE? I see /bin/startx (starts gnome3) /bin/startxfce4 (starts xfce4) Thanks for your help David

1 0

Re: [CentOS] Remote Desktop Recommendations (2)
by david 17 Nov '14

17 Nov '14

At 09:33 PM 11/16/2014, you wrote: >On Sun, Nov 16, 2014 at 9:59 PM, david <david(a)daku.org> wrote: > > > > I'm looking for a way to access a GUI desktop on a Centos 7 server from a > > Windows client. I have tried: > > - X2go from EPEL (fails immediately, apparently they are working on a > > solution, but...) > >X2go works, just not with Gnome3. Use KDE or MATE from epel. Also, >there was a bug in a recent x2goserver package that would fail if you >didn't have the x2goserver-xsession package installed. The fix for >that was done but I don't know if it ever got out of epel-testing. > >-- > Les Mikesell > lesmikesell(a)gmail.com Thanks, Les, for that detail. I knew that x2go had a known bug, but didn't know the name of the missing package. I explicitly installed it, but I still get failure asking for the MATE desktop. To summarize, my client shows the X2go screen, then it goes away. On the server, I see in /var/log/messages, a sequence such as: ------------------------ Nov 17 08:58:07 goat systemd: Created slice user-1000.slice. Nov 17 08:58:07 goat systemd: Starting Session 4 of user UUUUU. Nov 17 08:58:07 goat systemd: Started Session 4 of user UUUUU. Nov 17 08:58:07 goat systemd-logind: New session 4 of user UUUUU. Nov 17 08:58:11 goat /usr/bin/x2gostartagent: successfully started X2Go agent session with ID UUUUU-50-1416243489_stDMATE _dp32 Nov 17 08:58:15 goat kernel: fuse init (API version 7.22) Nov 17 08:58:15 goat systemd: Mounting FUSE Control File System... Nov 17 08:58:15 goat systemd: Mounted FUSE Control File System. Nov 17 08:58:15 goat /usr/bin/x2gosetkeyboard: Setting X keyboard according to /home/UUUUU/.x2go/C-UUUUU-50-1416243489_stD MATE_dp32/keyboard Nov 17 08:58:16 goat /usr/bin/x2goruncommand: launching session with Xsession-x2go mechanism, using STARTUP="mate-sessio n" Nov 17 08:58:16 goat /usr/bin/x2gomountdirs[3210]: successfully mounted VVVVV@127.0.0.1:30005/cygdrive/C/Users/VVVVV/X2GO~ 1/S-743F~1/spool to /tmp/.x2go-UUUUU/spool/C-UUUUU-50-1416243489_stDMATE_dp32 Nov 17 08:58:18 goat /usr/bin/x2goumount-session[3574]: successfully unmounted "/tmp/.x2go-UUUUU/spool/C-UUUUU-50-14162434 89_stDMATE_dp32" Nov 17 08:58:28 goat systemd-logind: Removed session 4. ------------------------- In the above, UUUUU is the user name on Centos 7, and VVVVV is the name on Windows, as evidenced by the reference to 'cygdrive'. --- In response to others that suggested VNC, I used to use it, but was disappointed with the speed when used remotely. On an internal LAN connect, it worked fine, but otherwise it became painful to use. NX became my next choice for Centos 6, but the x2go package became my favorite because of the ease of installation -- (nx needed lots of other things at setup, and appears unavailable for C7). David

3 2

Re: [CentOS] automated stopserver & shutdown?
by James B. Byrne 17 Nov '14

17 Nov '14

On Mon, November 17, 2014 01:17, Hadi Motamedi wrote: >> A simple script which loops though a text file that lists the hostname / >> username / password for the login and then runs the "shutdown - h now" >> command? Though, I would be uneasy with listing root passwords in a >> random text file. >> > > Thank you for your help . So how to open a session with another server > to login with root password and issue a command on it ? > > I do not believe that, on reflection, you really would want to do that. An alternative approach is to use ssh-keygen to create a key pair for root on the control host and then add that public key to the authorized_keys file inside /root/.ssh on each of the target hosts. So, something like this: On control host as root user: ssh-keygen -t rsa -b 4096 Generating public/private rsa key pair. Enter file in which to save the key (/<path to <<userid>> home directory>/.ssh/id_rsa): Enter passphrase (empty for no passphrase): Enter same passphrase again: Your identification has been saved in /<path to <<userid>> home directory>/.ssh/id_rsa. Your public key has been saved in /<path to <<userid>> home directory>/.ssh/id_rsa.pub. The key fingerprint is: 1c:6c:91:76:64:83:a5:32:4e:ac:df:28:ed:cb:9a:dd <userid>@<hostname> That will give you the following files in /root/.ssh ll .ssh total 28 -rw-r--r--. 1 root root 4619 Nov 17 2014 authorized_keys -rw-------. 1 root root 3239 Nov 17 2014 id_rsa -rw-r--r--. 1 root root 756 Nov 17 2014 id_rsa.pub Now login to each of your target hosts as root and transfer /root/.ssh/id_rsa.pub from the control host and append (>>) it to /root/.ssh/authorized_keys on the target host. If the .ssh directory does not already exist on the target host then you should first create the necessary files by running ssh-keygen on the target (otherwise you need consider whether or not selinux is enforced or not? is the mode set correctly? yada, yada). Once you have setup the target hosts for ssh and have added the control host's public key to .ssh/authorized_keys then you should be able to ssh into each from the control host without having to provide a password. You will however need to add each remote host's identity (/etc/ssh/ssh_host_rsa_key.pub) to your own /root/.ssh/known_hosts file the first time that you connect. This can be automatically created and maintained by the ssh client so long as the remote host's ssh identity is not subsequently changed. If you have previously ssh'ed into the remote hosts than this will already have been done. Once the connections have been set up between the control host and all of the target hosts then you should be a able to do this (mind the -t option to ssh): for host in \ long_host_name.domain.tld \ targethost{01,02,03,04,05,06}.domain.tld \ othername{x,y,z}.domain.tld \ yet_another_very_long_host_name.domain.tld \ 192.168.0.2{1,5,7,8,9} ; do ssh -t $host 'echo -e "\n\n$HOSTNAME\n" ; /root/stopServer ; init 0'; done Personally, I would use a separate account for this and add that account to the /etc/sudoers file on each of the targets: shutdown_userid NOPASSWD: localhost=/sbin/init 0 shutdown_userid NOPASSWD: localhost=/sbin/shutdown -h now shutdown_userid NOPASSWD: localhost=/root/stopServer The ssh-keygen / authorized_hosts setup would then need to be done in that account's home directory on each target rather than root's. Then use 'sudo /root/stopServer ; sudo init 0' to run the shutdown scripts. I also would use 'shutdown -h now' instead of 'init 0' -- *** E-Mail is NOT a SECURE channel *** James B. Byrne mailto:ByrneJB@Harte-Lyne.ca Harte & Lyne Limited http://www.harte-lyne.ca 9 Brockley Drive vox: +1 905 561 1241 Hamilton, Ontario fax: +1 905 561 0757 Canada L8E 3C3

1 0

EL5 Security Policy for the final 3 years
by Johnny Hughes 17 Nov '14

17 Nov '14

Red Hat's Security policy for Production 3 Phase of the Life Cycle for EL5 is that they will only release "Critical impact Security Advisories (RHSAs) and selected Urgent Priority Bug Fix Advisories (RHBAs) may be released as they become available. Other errata advisories may be delivered as appropriate." https://access.redhat.com/support/policy/updates/errata/#Production_3_Phase In practice, what that means so far is this: All Important and Critical security updates have been released for EL5, but some moderate and below security updates have not been, and are not going to be released by Red Hat for EL5. I do not agree with this policy, but it is not one that the CentOS Project (or I) have any say about. These updates will not be released for RHEL-5 ... therefore they will also not be released for CentOS-5. Due to this security policy, I highly recommend moving CentOS-5 based workloads to CentOS-6 and that every user stop using CentOS-5 as soon as possible. Here is a list of updates that are not done on RHEL-5 and are not planned to be done at this time by Red Hat for RHEL-5 (and therefore CentOS-5): > ruby Moderate https://access.redhat.com/security/cve/CVE-2014-8080 > python Low https://access.redhat.com/security/cve/CVE-2014-7185 > libgcrypt Moderate https://access.redhat.com/security/cve/CVE-2014-5270 > wget Moderate https://access.redhat.com/security/cve/CVE-2014-4877 > perl-Data-Dumper Low https://access.redhat.com/security/cve/CVE-2014-4330 > cups Moderate https://access.redhat.com/security/cve/CVE-2014-3537 > dbus Moderate https://access.redhat.com/security/cve/CVE-2014-3477 > dovecot Moderate https://access.redhat.com/security/cve/CVE-2014-3430 > exim Low https://access.redhat.com/security/cve/CVE-2014-2972 > cups Moderate https://access.redhat.com/security/cve/CVE-2014-2856 > openssh Moderate https://access.redhat.com/security/cve/CVE-2014-2653 > libxml2 Moderate https://access.redhat.com/security/cve/CVE-2014-0191 > qemu Moderate https://access.redhat.com/security/cve/CVE-2013-6458 > squid Moderate https://access.redhat.com/security/cve/CVE-2012-5643 > openssh Low https://access.redhat.com/security/cve/CVE-2014-2532 > libX11 Moderate https://access.redhat.com/security/cve/CVE-2013-1997 > libFS Moderate https://access.redhat.com/security/cve/CVE-2013-1996 > libXext Moderate https://access.redhat.com/security/cve/CVE-2013-1982 > I wish there was another option, but I just don't see any others .. I know I would not use packages with moderate security issues unfixed in production on purpose. I think this is a ridiculous policy, but it is what it is. Thanks, Johnny Hughes

4 3

← Newer
1
...
5
6
7
8
9
10
11
...
20
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Discuss November 2014