Discuss

discuss@lists.centos.org

34753 discussions

Re: [CentOS] CentOS-6.7, kvm bridges, virtual interfaces, and routes
by James B. Byrne 11 Jan '16

11 Jan '16

On Sat, January 9, 2016 19:48, Gordon Messmer wrote: > On 01/09/2016 03:30 PM, isdtor wrote: >> Search for policy routing. > > Policy routing isn't relevant. > > In order to communicate across a LAN, two hosts must be in the same > broadcast domain. Hosts in 192.168.51.0/24 cannot communicate with > hosts in 192.168.52.0/24. > > If I have all of the kvm guests on both hosts, together with the br0 bridge on both hosts, configured with addresses on the same a.b.c.0/24 network then will all communication on a.b.c.0/24 pass over br0 if the target address is on the other host? kvmh1g1 eth0=192.168.51.100 kvmh1 br0=192.168.51.41 kvmh2 br0=192.168.51.42 kvmh2g1 eth0=192.168.51.200 In other words, with the address configuration given above, will traffic from 192.168.51.200 reach 192.168.51.100 via the cross-over cable between 192.168.51.42/192.168.51.41? -- *** e-Mail is NOT a SECURE channel *** Do NOT transmit sensitive data via e-Mail 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

2 1

Learned something today
by Alice Wonder 11 Jan '16

11 Jan '16

Did you know you can use systemd to turn on tmpfs for /tmp ??? I had no clue that was possible, always did it through fstab - but systemctl is-enabled tmp.mount That will tell you if systemd has configured /tmp to mount as tmpfs (default in CentOS 7 is no) systemctl enable tmp.mount That, as you can probably guess, tells systemd to mount /tmp as tmpfs next time the system boots. -=- Just thought I would pass it along in case others may benefit. For me, I only need /tmp as tmpfs on my Bitcoin box - and then only when generating private keys for cold storage, SSDs are often not very good at securely deleting files. So I use tmpfs for /tmp and generate the private keys for cold storage to a file in /tmp and then print it from there (for storage in safe deposit box) - so that the private keys can't be recovered from the SSD. Yes, booting off a CD to generate and then print works too, but that's bothersome. Anyway just in case anyone else needs /tmp as tmpfs for anything, thought I'd pass the tip along.

4 3

Centos 3.8 Server Questions, SeaMonkey Mozilla and Java
by H 11 Jan '16

11 Jan '16

In order to run a certain software package that runs as a java applet I had to install Centos 3.8 on a 32-bit server. After installation I upgraded the installation using yum after repointing the configuration file to vault.centos.org. This worked fine, however, I still have to resolve two problems: - I'd would like to make the EPEL repository available but have not been able to find if old EPEL software packages are stored somewhere else for non-supported versions of CentOS? - I am trying to get Java 1.4.1 running in SeaMonkey Mozilla 5.0, the latest version I have been able to find for Centos 3.8, but have not had success to date. My current understanding is that I need to create a symbolic link to a Java .so file in the Mozilla plugin directory but must be doing something wrong since I cannot get Java to show up as a plugin using about:plugins in SeaMonkey Mozilla. Is anyone able to offer suggestions? Thank you. H

9 13

How to correct LiveKDE stick?
by Timothy Murphy 11 Jan '16

11 Jan '16

CentOS-7-x86_64-LiveKDE-1511.iso crashes on my AMD/ATI Radeon machine. I installed CentOS-7.2 by first installing CentOS-7-x86_64-LiveKDE-1503.iso, then appending GRUB_CMDLINE_LINUX_DEFAULT="initcall_blacklist=clocksource_done_booting" to /etc/default/grub and running update-grub. My question is: would there be any way, short of re-compiling the ISO, of altering the grub.cfg seen when booting from a USB stick? -- Timothy Murphy gayleard /at/ eircom.net School of Mathematics, Trinity College, Dublin

2 1

MATE question
by Fred Smith 10 Jan '16

10 Jan '16

I know I should go to EPEL with this, but thought to ask here in case anyone else uses Mate and may know more about it than I do. I had installed the MATE 1.12 packages from epel testing a few days ago, but since that did not include some of the devel packages, I just removed it and reinstalled all the MATE 1.10 packages from epel-7. they all installed fine, but now my desktop has a background, and top and bottom panels and nothing else. right-click on the desktop produces no context menu. Anyone know what I've missed (or messed) ? thanks! -- ------------------------------------------------------------------------------- Under no circumstances will I ever purchase anything offered to me as the result of an unsolicited e-mail message. Nor will I forward chain letters, petitions, mass mailings, or virus warnings to large numbers of others. This is my contribution to the survival of the online community. --Roger Ebert, December, 1996 ----------------------------- The Boulder Pledge -----------------------------

4 8

CentOS-6.7, kvm bridges, virtual interfaces, and routes
by James B. Byrne 10 Jan '16

10 Jan '16

I have been looking at this problem on and off for a considerable period. Given my lack of knowledge I have been unable to resolve this quickly and in consequence it has been constantly shoved to the background as other issues arise. Here is the situation: I have two dual-homed kvm hosts both running CentOS-6.7 and identically configured. These are connected to the same LAN segment via br0/eth0 and to each other via a cross-over cable on br1/eth1. The IPv4 assigned to br0 on both is a publicly routeable address. The IPv4 assigned to br1 on both is a private address in the 192.168.0.0/16 address space. The guests on each host have their virtual eth0 bridged with their host's br0 and eth1 bridged with their host's br1. The addresses used by the guests on eth0 are publicly routeable, the addresses used on eth1 are private. I would like to configure br1/eth1 on both kvm hosts such that each is a gateway to the other. I then also would like to configure each kvm guest of each host such that their traffic to the private network segment on the opposite host routes through the x-over cable via br0 whilst everything else goes out through br1 to the LAN and gateway. Has anyone here done anything like this? If so, can you point me to any online resource that could more or less walk me through the process without me having to complete the coursework for a network engineer. I just want to keep data transfer traffic between pairs of kvm guests off of the public lan without having to install more hardware. The existing cabinets are not going to support it either space wise or power wise. An ASCII art diagram might help, or might not. <pre> kvmh1g1 eth0/192.168.51.1 eth1/aaa.bbb.ccc.151 <-------------> | | kvmh1 br1/aaa.bbb.ccc.51 | |---> br0/192.168.51.1 | X | kvmh2 |---> br0/192.168.52.1 | br1/aaa.bbb.ccc.52 | | kvmh2g1 eth0/192.168.52.1 | eth1/aaa.bbb.ccc.251 <-------------> | | gateway eth1/aaa.bbb.ccc.1 <---------------> | </pre> I have tried multiple approaches without success and in so many variations that I no long can clearly recall the details. At the moment my thought is that if br0 was set to 192.168.51.1/24 on kvmh1 and to 192.168.52.1/24 on kvmh2 and a routing table entry was made on kvmh1 to send traffic addressed to 192.168.52.0/24 through 192.168.51.1/24. And on kvmh2 br0 was set to 192.168.52.1/24 and a routing table entry was made on kvmh2 to route traffic to 192.168.51.0/24 through 192.168.52.1/24. I thought that if the kvm virtual guests on kvmh1 were then configured to use addresses from 192.168.51.0/24 while those on kvmh2 used 192.168.52.0/24. And everything was configured to use their respective host's br1 address as their gateway then this should work. But I am evidently either fundamentally wrong or I have misconfigured things somehow. Should this set up work as I imaging? What would be the correct static routing table entries to make it work? -- *** e-Mail is NOT a SECURE channel *** Do NOT transmit sensitive data via e-Mail 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

4 3

dhcpd and centos 7 network burps
by david 09 Jan '16

09 Jan '16

Folks I have noticed that my Centos 7 systems which use DHCPD from a router or gateway all seem to fail when the DHCP client asks for a renewal. After issuing systemctl restart NetworkManger the connectivity resumes until the lease renewal time occurs again. As part of my research, I set up a CRON job that runs every five minutes, and uses ip route to determine what my IP address is. When it fails to get an address, it issues the restart NetworkManager. This problem seems to occur on various DHCPD servers, but the 30-minute regularity is surprising. Is there an issue with the DHClient and the renewal process whereby it looses connectivity? Centos 6 and Windows have no problem using the same DHCP server. David

1 0

Yum Weird Message
by Alice Wonder 08 Jan '16

08 Jan '16

I ran into this exact issue last night - http://www.iotti.biz/?p=433 When a computer is connected via IPv4 but the IPv4 a repo host connects to is not available, yum then tries the IPv6 address and will fail with a confusing message telling you it failed to connect to the IPv6 address. I don't know if there is a way for yum to figure out whether the current network connection to the Internet is IPv4 or IPv6. But if there is a way, it might make a usability improvement. A lot of people have no idea what IPv6 is and would be confused. I was confused myself at first, wondering if DHCP pulled in IPv6 from the router. -- -=- Sent my from my laptop, may not be able to respond timely

4 7

yum-cron / email sending problem
by Joey 08 Jan '16

08 Jan '16

I want to use yum-cron with email notificaton. The upgrade runs, but i get no message. yum-cron aborted with this message: [root@h1 yum]# yum-cron Traceback (most recent call last): File "/usr/sbin/yum-cron", line 711, in <module> main() File "/usr/sbin/yum-cron", line 708, in main base.updatesCheck() File "/usr/sbin/yum-cron", line 631, in updatesCheck self.installUpdates(self.opts.update_messages) File "/usr/sbin/yum-cron", line 582, in installUpdates self.emitMessages() File "/usr/sbin/yum-cron", line 694, in emitMessages map(lambda x: x.sendMessages(), self.emitters) File "/usr/sbin/yum-cron", line 694, in <lambda> map(lambda x: x.sendMessages(), self.emitters) File "/usr/sbin/yum-cron", line 227, in sendMessages msg = MIMEText(''.join(self.output)) File "/usr/lib64/python2.7/email/mime/text.py", line 30, in __init__ self.set_payload(_text, _charset) File "/usr/lib64/python2.7/email/message.py", line 226, in set_payload self.set_charset(charset) File "/usr/lib64/python2.7/email/message.py", line 262, in set_charset self._payload = self._payload.encode(charset.output_charset) UnicodeEncodeError: 'ascii' codec can't encode characters in position 209-210: ordinal not in range(128) whats happend please?

1 0

SELinux context change on /etc/posfix/main.cf
by James B. Byrne 08 Jan '16

08 Jan '16

This morning I received this report of a change to the SELinux context of /etc/posfix/main.cf on one of our hosts. from: system_u:object_r:postfix_etc_t:s0 to: unconfined_u:object_r:postfix_etc_t:s0 The contents of the file have been verified as unchanged. There was a yum update applied yesterday to this host and this may be an intended alteration. However, can anyone confirm this for me? Or, otherwise explain what has happened? -- *** e-Mail is NOT a SECURE channel *** Do NOT transmit sensitive data via e-Mail 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

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