There is a difference between the DHCP server and DHCP client. The DHCP server should have no play in this. What does is the DHCP client. In the messages log this is all the dhclient messages.

One or two people have had issues like this and we have have not been able to fathom them out. You may be the second or third person. I am not sure. If you can switch to static, it should sidestep the issue but it is perplexing.

I got it to keep the gateway overnight

Here is what I noticed DHCP server was running but no interface was enabled on it so I uninstalled it. Not sure if that had any impact

Enabled all interfaces set both to static

Set the one for internet access to external and the other that was not connected to lan

Gave both an ip address on different subnets (local ones) even the one that was not connected.

I had both set to static before and one set to external and one set to lan. However this time I gave both of them an IP and other info. instead of just one.

Keeping my fingers crossed

Please also check that syswatch is running.

Nick Howitt wrote:

There is no harm setting it.

Can you look in /var/log/syswatch and /var/log/messages to see if there are any lease renewal issues.

I went back to static

Removed the network device I was not using to see if that helps

If it looses gateway again, I have no clue what is going on

There is no harm setting it.

Can you look in /var/log/syswatch and /var/log/messages to see if there are any lease renewal issues.

Nick Howitt wrote:

I'm confused. You should not be running network manager at all. The standard ClearOS installation does not use it and I believe it is incompatible with syswatch. Normally that parameter does not appear in any of the ifcfg-* files.

Well I guess that will not fix it...

I found two articles talking about that setting thought it might help.

I'm confused. You should not be running network manager at all. The standard ClearOS installation does not use it and I believe it is incompatible with syswatch. Normally that parameter does not appear in any of the ifcfg-* files.

HP ProLiant DL360e Gen8 Server - No Network Connectivity After Installing CentOS 6.3
Issue
CentOS 6.3 installed on HP ProLiant DL360e Gen8. After installing there was no network connectivity. No LEDs on the network interface. Attempted to load the network driver and restart the server with no success.

By default CentOS 6 will not configure network interfaces on a new installation which means the host will not have network connnectivity.

File configurations for each network device are located in the /etc/sysconfig/network-scripts/ifcfg-{eth0,eth1,bond,etc} directory. To modify the network address manually, edit the file ifcfg-ethN.

Solution
Check the contents of the ifcfg-eth0 file:

#cat /etc/sysconfig/network-scripts/ifcfg-eth0
DEVICE="eth0"
HWADDR="MAC Address"
NM_CONTROLLED="yes"
ONBOOT="no"

The above indicates that eth0 is disabled. With a text editor of ones choice, open the /etc/sysconfig/network-scripts/ifcfg-eth0 and edit as indicated below:

#vi /etc/sysconfig/network-scripts/ifcfg-eth0
DEVICE="eth0"
HWADDR=MAC Address*System MAC*
NM_CONTROLLED="no"
ONBOOT="yes"
BOOTPROTO="dhcp"

DHCP Configuration

DEVICE="eth0" – Name of Network Interface.

HWADDR=MAC Address*System MAC* – MAC address of Network Interface.

NM_CONTROLLED="no" – Network Manager should disable since it is unnecessary in this configuration.

ONBOOT="yes" – Means the NIC will be started during boot.

BOOTPROTO="dhcp" – Instruct the OS to find a DHCP server from which to obtain an IP address. The DHCP server will assign all necessary network information including IP address, Netmask, Gateway and DNS server.

Restart the network service

#service network restart
For static IP Configuration just changed the bootproto to static (BOOTPROTO="static"), specifify the IP, Netmask and restart the service as well.

Found this on the interwebs

to /etc/sysconfig/network-scripts/ifcfg-eth0, and either do service network restart, or reboot. And make really sure NetworkManager's disabled by changing NM_CONTROLLED="yes" to NM_CONTROLLED="no" in the same file, that thing's a vile horror for servers.

so ifcfg-eno1 added NM_CONTROLLED="no" HPs website also mentioned to turn this off

Crossing fingers

I have two interfaces on this machine

en0
en1

I have removed en1 to see if that helps

en0 is dhcp external

After making changes I rebooted

thankfully I can use the google cloud host as a backup as I get this one sorted out

Nick Howitt wrote:

If you do not have a dedault route, it generally means you do not have an external interface configured. Can you check that eno1 is set to External and not LAN?

Yes it is set to external dhcp

I had this same issue with google cloud at one point. Setting it to external worked

However on this host it does not seem to resolve it.

If I go to the network screen in centos and do update

The default gateway reappears and ever thing starts working.

However overnight it seems to loose it.

If you do not have a dedault route, it generally means you do not have an external interface configured. Can you check that eno1 is set to External and not LAN?

Well I woke up this morning and this

I can not connect to the internet

[root@vonwallace ~]# route -n
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
192.168.86.0 0.0.0.0 255.255.255.0 U 0 0 0 eno1
[root@vonwallace ~]#

Why does it keep doing this?

No clue...

[root@vonwallace ~]# ifconfig
eno1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.86.43 netmask 255.255.255.0 broadcast 192.168.86.255
inet6 fe80::c634:6bff:fe93:6ccc prefixlen 64 scopeid 0x20<link>
ether c4:34:6b:93:6c:cc txqueuelen 1000 (Ethernet)
RX packets 168601 bytes 150315576 (143.3 MiB)
RX errors 0 dropped 1 overruns 0 frame 0
TX packets 55040 bytes 13022886 (12.4 MiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
device interrupt 16

eno2: flags=4099<UP,BROADCAST,MULTICAST> mtu 1500
ether c4:34:6b:93:6c:cd txqueuelen 1000 (Ethernet)
RX packets 0 bytes 0 (0.0 B)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 0 bytes 0 (0.0 B)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
device interrupt 17

lo: flags=73<UP,LOOPBACK,RUNNING> mtu 65536
inet 127.0.0.1 netmask 255.0.0.0
inet6 ::1 prefixlen 128 scopeid 0x10<host>
loop txqueuelen 1000 (Local Loopback)
RX packets 15518 bytes 2636932 (2.5 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 15518 bytes 2636932 (2.5 MiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

[root@vonwallace ~]#

Yes, the graphical screen only gets the Network screen.
I don't understand it, but some systems go straight to the graphical screen when logging on at the console and others don't and you have to select it from the small menu.

/etc/clearos/network.conf

Did find some left overs in there

Cleaned them up everything else looked good

Now that both en interface are setup as External and DHCP the gateway is not lost. one nic is not attached to a cable

Not sure why static external had an issue with loosing gateway.

Another question when at the console on the server its self logging in via the graphical menu not the text one. Should I only get the network config screen of the graphical menu?