1

In my structure I have a VLAN called 172, which is exclusive to communication between the iSCSI storage and the blades of the server. The VLAN is configured in two Dell PowerConnect M8024 switches, which are connected to the blades. The ports are in trunk mode.

I have RHEL 7.6 installed in 2 Dell PowerEdge blades, where I configure the p3p1 and p3p2 interfaces as follow:

/etc/sysconfig/network-scripts/p3p1
DEVICE=p3p1
TYPE=Ethernet
BOOTPROTO=none
ONBOOT=yes

/etc/sysconfig/network-scripts/p3p2
DEVICE=p3p2
TYPE=Ethernet
BOOTPROTO=none
ONBOOT=yes

The VLAN are defined in the following files:

File: /etc/sysconfig/network-scripts/ifcfg-p3p1.172
DEVICE=p3p1.172
BOOTPROTO=none
ONBOOT=yes
IPADDR=172.16.0.38
NETMASK=255.255.255.0
GATEWAY=172.16.0.254
VLAN=yes

File: /etc/sysconfig/network-scripts/ifcfg-p3p2.172
DEVICE=p3p2.172
BOOTPROTO=none
ONBOOT=yes
IPADDR=172.16.0.38
NETMASK=255.255.255.0
GATEWAY=172.16.0.254
VLAN=yes

After I edit those files and restart the network service with the command

systemctl restart network.service

the VLAN tagging works and I'm able to ping other hosts in the network, but if I restart the machine, the VLAN tagging does not work anymore, becoming impossible to ping other hosts . If I check the files and the IFs on 'nmtui' they all there and active, correctly configured.

Some files get additional info after sometime, but I guess it's normal:

FILE: /etc/sysconfig/network-scripts/ifcfg-p3p1.172
DEVICE=p3p1.172
BOOTPROTO=none
ONBOOT=yes
IPADDR=172.16.0.38
NETMASK=255.255.255.0
GATEWAY=172.16.0.254
VLAN=yes
TYPE=Vlan
PHYSDEV=p3p1
VLAN_ID=172
REORDER_HDR=yes
GVRP=no
MVRP=no
PROXY_METHOD=none
BROWSER_ONLY=no
PREFIX=24
DEFROUTE=yes
IPV4_FAILURE_FATAL=no
IPV6INIT=no
NAME="Vlan p3p1.172"
UUID=fb214dde-368b-3d9c-9c1a-e8f1c5e9deb7

The only way to make it work again is deleting the VLAN IF files (p3p1.172 and p3p2.172), restart the network service, then recreate those files and restart the network service. Again, if I reboot, it stops working.

The switch config is OK as far as I can see.

Anyone has any idea why this happens? Is there other files that might get overwritten during restart?

Thanks in advance.

  • 1
    Is this the same IP on both interface ? you also ned bonding – Archemar Apr 4 at 16:39
  • Did you explicitly disable NetworkManager and enable the legacy network service? – Michael Hampton Apr 4 at 16:57
  • @Archemar Yes, is the same IP on both interfaces on each blade. I will check abou net bonding, I didn't configured that. – Marco Antônio Apr 4 at 21:37
  • @MichaelHampton no, I didn't disable it. Might be this that's causing part of the mess? – Marco Antônio Apr 4 at 21:38
  • @MichaelHampton is NetworkManager better than the legacy service? – Marco Antônio Apr 4 at 21:41
0

Thanks to the tips from @MichaelHampton and @Archemar, I deleted the faulty network scripts and redid the config using only the NetworkManager, via 'nm-connection-editor' GUI. I've created the bond and vlan through there and it persisted through reboot.

After those tips I´ve read in the RedHat documentation that the NetworkManager Daemon will have priority over the legacy network.service interfaces, totally ignoring them and causing the described problem.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.