NS7a3 - transaction check error on conflicting packages

Development Testing
,
1

After installing NS7 alpha3 on a VM, either in interactive or unattended mode, the network interface is assigned an IP but cannot connect to Internet nor ping to it. Due to that (I guess), yum update was getting:

Cannot find a valid baseurl for repo: base/7/x86_64

Network worked after an additional reboot. Maybe it’s because of the VM…

yum update is reporting some conflicts:

Running transaction check
Running transaction test

Transaction check error:
file /etc/e-smith/templates/etc/yum.conf/10base conflicts between attempted installs of nethserver-base-3.0.0-1.ns7.noarch and nethserver-yum-1.3.6-1.1.g5515a5e.ns7.noarch
file /etc/e-smith/templates/etc/yum.conf/20proxy conflicts between attempted installs of nethserver-base-3.0.0-1.ns7.noarch and nethserver-yum-1.3.6-1.1.g5515a5e.ns7.noarch

Error Summary
--------------------

P.S. the activation of the feature reporting failed login attempts after user login is a nice touch.

2

Did you check the dns, it’s hard coded to google…

3

Just checked it right now. /etc/dnsmasq.conf holds google’s dns server, but it is not blocked on my gateway, and it worked after a reboot.

One of the test installs was in interactive mode using opendns servers and could not connect to Internet.

Maybe tomorrow can try to do another clean install to see if this behavior is consistent or just a temp issue.

4

I will check also this!

1 Like
5

Did a new unattended install (on virtualbox VM):

  • /etc/dnsmasq.conf has google’s dns and one dns from dhcp.
  • No connection to Internet or LAN. It works after bringing the interface down and up.
  • yum update failed as before.
6

Please wait the release of the latest nethserver-yum, to fix the RPM conflict!

You can download it from

http://packages.nethesis.it/nethserver/7.2.1511/base/x86_64/Packages/nethserver-yum-1.4.0-1.ns7.noarch.rpm

2 Likes
7

10h fix time? Wow

8

OK. will do. Thanks.

9

the new nethserver-yum rpm is working well :thumbsup:

2 Likes
10

Was having the same problem with NethServer 7 beta1 on virtualbox.
It happens when one of the network interfaces is attached to internal network.
Network manager leaves a stale dhclient process.

log showed errors like:

Jul 17 21:42:08 localhost systemd: Dependency failed for Network Manager Wait Online.
Jul 17 21:42:08 localhost systemd: Job NetworkManager-wait-online.service/start failed with result ‘dependency’. (…)
Jul 17 21:43:42 localhost network: Bringing up interface enp0s8:
Jul 17 21:43:42 localhost dhclient[3348]: dhclient(908) is already running - exiting. (…)
Jul 17 21:43:42 localhost network: [FAILED]
Jul 17 21:43:42 localhost network: RTNETLINK answers: File exists (…)
Jul 17 21:43:42 localhost esmith::event[3179]: Starting network (via systemctl): Job for network.service failed because the control process exited with error code. See “systemctl status network.service” and “journalctl -xe” for details. (…)
Jul 17 21:43:46 localhost esmith::event[984]: Event: system-init FAILED
Jul 17 21:43:46 localhost systemd: nethserver-system-init.service: main process exited, code=exited, status=1/FAILURE

it worked after doing one of these things:
a) bringing the interface down and up
b) killing dhclient
c) restarting network service
d) rebooting

1 Like
11

I agree: the network stack of CentOS 7 has some problems with dhclient especially when multiple interfaces use the DHCP.
We are investigating the problem and eventually we will open an upstream bug.

2 Likes