Ubuntu Networking :: NICs On Clean Server Installation Not Working?
Aug 18, 2010
I have a clean installation of Ubuntu Server 10.04 x64 on an HP Proliant DL380 G3. It has two Broadcom NICs in addition to its ILO. During the installation, both NICs were listed... but neither was able to pull DHCP, nor did they function with manual settings. The first NIC is currently connected and is known good. So I left it alone and figured I'd troubleshoot it later (which is now). This server is going to be a VMware Server host. I wanted to install ESXi on it, but it doesn't like the old server's cheap ICH RAID controller.
[Code]...
I noticed here it says they're disabled. so I thought perhaps this was a 64-bit driver issue. So, I booted my trusty Ubuntu desktop 10.04 32-bit live cd, and the NIC works fine. Now, here's where it gets weird. Seeing that, I booted it to Ubuntu server 10.04 32-bit installation from a flash drive, expecting to see the NICs working fine during the installation.
Just something that struck me while working on our virtual servers today.
I have bonded 3 NICs at the host in Ubuntu Server 8.04 LTS. They are using mode 0 for Round-robin. Point is to increase the speed/performance of all the servers, but mainly the fileserver. The fileserver is a virtual server running Ubuntu Server 8.04 LTS on VMware Server 2.0.
1) I noticed the NIC in the slave OS reported link speed as 1000 and Im unable to change it as the NIC (virtual one) doesnt support it. Does this not really matter, as the NIC doesnt exist, and it will run at higher speeds anyway? Or do I have to remove the bond on the host, bridge all 3 interfaces from the host to the slave OS, and then make a bond in the slave OS?
2) While at it, does mode 0 only increase performance on data being sent from the host or does it also increase the available incoming bandwidth?
I am trying to get preseed working on a bunch of machines with multiple NICs but it doesn't pick the right interface and/or gets "no link" on all interfaces. My PXE kernel line looks like so (I have auto=true priority=critical and interface=auto)
i have a Server, which has 2 nics installed. Each of those is connected to a router, which is connected to internet. On the server, i have apache, maillserver and im-server running. On the other hand, also squid, dansguardian and clam are running. so now: via eth0 i would like to have just the traffic, which is requested from outside (the big bad internet..) to my server (apache, mail, etc). via eth1 i would like to have all OUTGOING (also to the big bad internet) from the server, which is requested by a internal client. And of course all requests to my own server
both nics shall route their traffic to their own router. For better comprehension please consult the enclosed graph. Until now, i did not find a good solution, the default route is set to the traffic from eth0, if not, no external request will find back to a client do you have a idea how to handle this the easiest way?
configuring DHCP server with two NICs. I need DHCP server with 2 NICs:
eth1 - 192.168.103.11 eth2 - 192.168.123.11
The client also has two NICs and suppose to receive ip addresses eth1 from 192.168.103.0 and eth2 from 192.168.123.0 Here is the dhcpd.conf from dhcp server:
I recently set up a new web/file server with 9.10 server x64 with 2 NICs and I am trying to configure eth0 to respond to my LAN for internal samba filesharing and eth1 to handle website/ftp requests on my static IP, but whenever eth0 is up the server is not accessible at 173.XX.165.65 for web or ftp but both work fine at 10.1.10.100. When eth0 is down, public IP works fine. I have set /etc/network/interfaces like this:
Code: # The primary network interface auto eth0 iface eth0 inet static address 10.1.10.100
I have installed Ubuntu 64 10.04 server. I have two nics and have set them up to both be static with their own IP with the correct gateway, network, broadcast, subnet and dns-nameserver. When I have both enabled, I can ping local pc's but I can't ping Internet sites like Google nor can I get out to the Internet with apt-get or Lynx.
If I disable one, then I am able to get out to the Internet. All my configs look good, and it does not matter which one I disable, just so long as there is only one NIC on, everything is good.
I am trying to setup a HP blade (BL460) server with 2 nics on sles10sp3. I want one nic to connect to a management vlan and and second to connect to the standard network. It should be possible for a desktop in the standard network to connect to either nic by providing the correct ip address.use this server as our first virtualisation hosts server using Xen. I dont want traffic used to upload / download images to the server through nic1 to effect the users traffic on the standard network on nic2. However any attempt at routing (which Im not good at) has led to the tx traffic all going out the default route (users subnet).
Each time I setup the nics via yast2 lan I can get the standard lan nic working ok. but cannot get the management nic working correctly. The server can ping a workstation on another vlan, but the workstation cannot ping the management nic. The default route appears to be forcing all traffic on the host out through this route, which I presume is normal behaviour. Using tcpdump I can see the ping packets received by the server, the server then responds through the default gateway, which the workstation does not see.
However, this normal behaviour will result in extra traffic on the users network when image downloads are initiated from the management interface.If I use a 172.24.1.0 network routed through 172.24.1.1 router setup as our management vlan , and 200.200.1.0 with router 200.200.1.1 as our lan for general users. Where server ips are nic1 10.1.1.10 and nic2 200.200.1.10 My workstation would be 200.200.1.10. If I ping the server at 10.1.1.10 I get no response. If I ping the server at 200.200.1.20 no problem.
Basically I used the network setup recommended by vmware to manage a virtual server. I actually have 4 nics, I thought by just talking about 2 nics the problem would be easier to explain. vmware specify that two nics should be used for management and two for the Lan. This is what I am trying to achieve, but both subnets must operate independently.
Yesterday I upgraded to Ubuntu 10.10 by a clean install. My laptop is an acer aspire 5050 with a IEEE 802.11 b/g Wlan wireless card. I am using only Ubuntu in my laptop. Today I noticed that wireless is dead. The indicator is off and I can see that there is not detection of any wireless networks in the "Wireless Networks" menu. In this point I must inform you that I didn't experience any wireless problems with the two previous versions of Ubuntu (9.10 & 10.04). Below you can see the outputs (in italic format) of some commands that I peformed in terminal:
I have unexpected problem with yast2 in xen kernel!
My system is MB Asus p8h67-m evo CPU Intel i7 2600k cpu RAM DDR2 8gb
Ive install clean minimal server config (11.3 x64), make all updates from yast2, than reboot, than from yast2 start installation of XEN (in Virtualization, not in software management), reboot, set Xen dom0 to boot and than, after login, starting yast2 and got on blue screen with text "Loading modules,
sbin/yast2: line 399: 5427 Illegal instruction $ybindir/y2base $module "$@" "$SELECTED_GUI" $Y2_GEOMETRY $Y2UI_ARGS and return to promt
I have installed Ubuntu Server three times now. It "NEVER" asks for a "USERNAME". I can't log in to the clean install. Guess what? I never entered a username, so I don't know what it is. I have never logged in. Don't ask me to go to /etc/, I can't log in. It is NOT a video proplem, Ubunto Server is command line OS.
Login incorrect: [system-name] login: Login time out after 60 seconds.
I'm asked to enter a network system name during install. I'm asked for a password, then repeat it during install. but never asks for a "UserName" during install. So, what is it? root? ubuntu? default? Nope, I've tried many many many of them, even admin, nothing works.
I have a Dell PowerEdge 1650 server with 3 NICs in it. I am trying to use one NIC (eth0) to connect to the internet, the second NIC (eth1) to share the internet connection to a LAN, and the third NIC (eth2) to connect to the LAN.
It is running Ubuntu server 8.10
The problem that I'm having is that NetworkManager (version 0.7.0) keeps setting eth2 as the default connection, and then I don't get any internet at all from the server.
My main goal is to be able to share files from the server to computers on the LAN. The secondary goal is to have a virtual machine hosting a Halo server, connected through eth2, so that it can host LAN games. But that is a secondary goal after I get the connections to work.
I've been looking around and finding other people who have done bits and pieces of this, but not the whole thing. I had been hoping to simply use the NetworkManger, but I don't mind editing config files.
Also I can't seem to find any good instructions on editing the /etc/network/interfaces file. The man file is incomplete, and everything seems to point back to the man file.
I am trying to install CentOS 5.4 64-bit on a server with two NICs, one of which allows for iSCSI boot. I want the NIC that allows for iSCSI boot to connect to the iSCSI target on my 169.254 subnet. I want the other NIC to be connected to my 192.168.1 subnet so that it can reach the Internet for the CentOS Network Install.
Anaconda only configures one NIC, leaving me with an iSCSI target and no Internet or with Internet and no iSCSI target. Is it possible to configure both NICs in Anaconda?
Ok so I'm gonna try and explain this easily. I installed 11.04 32bit on an older HP P4 machine. Boots up and goes to classic desktop.. But networking looks like a wifi devices rather than a wired connection. I installed another NIC rather than the onboard but had the same result.. I also tried yet another PCI NIC and a USB NIC with the same result. I also tried to set my connection to manual and give myself a static IP, but had no such luck.
I know my internet connection works because I'm on another machine typing this to you all.. All the NICs also work in W7 on the dualboot setup I did.
Is it possible to install two NICs during an installation from CD? I tried and failed yesterday. The eth0 would have a static ip address 192.168.0.11 and would be used for Internet/network whereas eth1 would have a static ip address of 192.168.2.81 and would be used for connection to the ISCSI system (DROBOPRO). Oh yes, it's Ubuntu 10.04 Server version.
Install asks me to choose one of the two devices for configuration - I tried to be clever and set up one then the other but it didn't work, although the ISCSI subsystem was clearly working when I set up eth1 because the DROBO kicked into life as soon as I entered the IP address - communication was taking place. Thinking that my fudge had worked, I continued to complete the installation only to find that neither network interface worked on reboot.
I have attached a diagram to show the layout, but basically I have a server that has 2 nics. I use NIC1 for the server and its services. Im using VirtualBox to run VMs and want to use NIC2 for the VMs exclusively and I want the VMs on a different network completely from the host.Ive tried setting up NIC2 on the other network but I end up with networking problems on network1. Also I cant figure out how to tell VirtualBox to bind to the host NIC2.
I have 2 x PCs and a NAS. Both PCs have 2x NICS. PC connectivity to Internet is via an ADSL router. Current config: Thus far (by choice) I've used static IPs in the 192.168.168.x range for my internal network, connecting all PCs and NAS via a jumbo frame enabled gigabit switch. This has facilitated moving data between the PCs and the NAS at high-speed. As both PCs also required Internet access from time to time, both are also connected to the ADSL router using the 2nd NIC and using subnet 192.168.1.x. I'm sure some of you are shaking your heads by now, but it works well and has been entirely hassle free.
However, I've an app running on the NAS that I'm keen to get Internet connected also. As my existing network devices are not using DHCP I figured the simplest method would be to change my ADSL router configuration such that it is in the same 192.168.168.x subnet, change its DHCP server settings to serve IPs in the same subnet (but in a restricted range I know won't cause any conflicts with the static IPs) and problem solved. On changing the ADSL router confiruration with all machines already booted up and configured as described above, everything worked. All devices could see one another, and access the Internet. On later rebooting the system this no longer works
- Internet access is fine but PCs don't see one another or the NAS. If I disconnect the ADSL Router from the PCs then all devices see one another again.
- Does having 2x NICS on a single device each assigned unique IPs in the same subnet create an issue and can it be overcome? I'd like to overcome it because making one of my PCs the gateway forces me to have it on anytime another device needs access.
- If I'm forced to use Internet connection sharing with one PC on the network connected to the router, how do I best configure this?
- One of the things I need to retain is gigabit connectivity between the PCs and PCs and the NAS (currently achieved by 192.168.168.x subnet being linked via gigabit switch).
My brother has a Ubuntu server attached to a LAN with Windows PCs. I set up an openVPN tun service on the server, let's call this VPN1, so that I can connect remotely from my Ubuntu desktop. The server has one NIC and the LAN has a router that is the gateway to the internet.
My brother would like to remotely access his windows network when roaming with his Windows laptop.
Therefore, I would like to set up a second VPN service, this time tap, that is bridged with his LAN. Let's call this VPN2.
The LAN subnet is 192.168.1.0/24. The internet gateway is 192.168.1.1
The NIC has a fixed IP set by the router of 192.168.1.150
How do I set up VPN2 so as not to trash VPN1? That is, if I bridge eth0 with tap0 how will that impact VPN1's tun0 that is using eth0? Do I need a separate NIC for the VPN2 bridge?
I am considering moving to a software (ebox/pfSense/Zeroshell etc) based router.
It fits well to have 1 NIC connecting to your modem & the other to your switch & network. It seems to make sense it is more secure, but is it?
If I assign the 1 NIC in my router 2 IPs, a subnet for the modem and a subnet for the LAN, how is the different?
I can't think exactly what it is. I am no network guru and probably missing something down at a lower networking level, but for a home router, does 1 or 2 NICs make any difference?
I have a pc with debian 6 (without GUI) installed on it and want to use it as server at home. It has 2 ethernet nics. Now i want to configure the routing process. Searched internet for a long time found something but couldn't get it work.
My server has 2 network cards. I have SLES 10 SP2 installed on this server. I had configured Ethernet bonding on this machine using 2 NICs. Now I want to assign 2 different IP's to 2 different NICs. I am able to do this without any issue using yast. ifconfig command shows that eth0 has 192.168.1.10 and eth1 has 192.168.1.12. Rest all networking details are same (netmask, gateway, DNS)
Strange thing: When I unplug only eth0, (eth1 is plugged in correctly) I am not able to ping any IP. If I unplug only eth1, (eth0 is plugged in) I am able to ping both IP's.
I am currently investigating channel bonding for rhel5.5.Is there any way I can create one bond from nics on two different linux machines.They will be a cluster without any clustering software like beowulf or LVS.There will be a cluster interconnect through which they can communicate.Should I tweak the bonding module??
I'm running Ubuntu 11.04 and I'm really new to linux. My problem is that whenever I try to browse a site I notice the website loads very slowly because it takes a long time to do lookups. I installed Ubuntu with an onboard NIC and later switched to a PCI NIC (Dlink DGE-530T). Although I disabled the onboard NIC in the BIOS, it doesn't help. Could this conflict in configuration be a problem? My download rates are fine, its just lookups that take really long ( upto ~ 10 seconds). I know the PCI network card is fine because when I jump to Windows 7, lookups are normal again (~ 300ms). At first I thought about installing the sk98lin drivers for the PCI NIC but I saw a couple of places where people have mentioned that the skge driver that comes along with the kernel is better.
I have tried a system wide as well as Firefox disable of IPv6. Here is my /etc/udev/rules.d/70-persistent-net.rules Code: # PCI device 0x1186:0x4b01 (skge) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:24:01:14:eb:39", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth0" # PCI device 0x10de:0x0373 (forcedeth) SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:1e:8c:3e:19:ed", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth1" The interface I want to use according to the listing above is the one with the MAC - 00:24:01:14:eb:39.
I tried removing one of the entries in the file above and rebooting but it still didn't work. Here is a look at my /etc/network/interfaces Code: auto lo iface lo inet loopback # The primary network interface auto eth0 iface eth0 inet static address 192.168.1.10 gateway 192.168.1.1 netmask 255.255.255.0 network 192.168.1.0 broadcast 192.168.1.255
I have a MB with 2 NICs and on 1 I would like to have a static IP, on the other a DHCP for a local network.The problem is the DHCP provided gateway is made default vs the one for the static IP and resolv.conf is overwritten.Any suggestions how could this be resolved? I can tweak both the DHCP server settings and the box settings.
I have two NIC's interfaces on my linux machine(eth1 and eth2). Each have different IP addresses(10.0.0.1, 10.67.7.1). These two interfaces are connected together through hub. Here is the my question?
1) If I 'ping 10.0.0.1', it should go out through network interface eth2 and through hub and enter on eth1 and response also travel through similar direction.
2) If I 'ping 10.67.7.1', it should go out through network interface eth1 and through hub and enter on eth2.
How can setup routing table for this,I have tried setting up routing and iptables, etc.. nothing helped.
If any one good router/networking guy, you should know this one.I am doing a project, I want this way to handle this.
I'm trying to work out how to route all traffic destined for the internet from all devices connected to eth0 to a wireless router access point via wlan0 on my Slackware box. I also have dhcpd providing ip addresses on the same subnet to any device connected to both eth0 and wlan0.
If I connect to the router/access point via wireless or directly to the Server via a crossover cable I can obtain an ip address from dhcpd, so that works. As far as I can see I just need to how to route between eth0 and wlan0 then I can provide internet access to those devices!
ifconfig:
Code:
eth0 Link encap:Ethernet HWaddr **:**:**:**:**:** inet addr:192.168.2.253 Bcast:192.168.2.255 Mask:255.255.255.0 inet6 addr: fe80::201:2eff:fe27:aea3/64 Scope:Link UP BROADCAST MULTICAST MTU:1500 Metric:1
I currently have Samba working properly using the eth0 interface on my notebook computer. More importantly, I can browse the network and see the shares set up in Samba. If I disconnect and go wireless using the wlan0 interface I can no longer browse the shares on Samba via name. I can however browse via IP address of the wlan0 interface. How can I make it so that Samba is accessible by name when I switch between interfaces?
Code:
[global] workgroup = startrek server string = NCC-1701 Enterprise bind interfaces only = yes
I'm running 2 asterisk box (Centos) ,and the two servers are connected through 2 iax trunks.on one server i have 2 nics ,each one is connected to internet,now i want to use isp 1 for the first trunk and isp 2 for the second trunk.
iax 1 is listening on 12345 iax 2 is listening on 11223
is there a way to split traffic on both nic using ports (netfilter iptables)?
