Networking :: Bonded Ethernet Interface Can Ping Internally But Not Externally
Oct 14, 2009
So I keep reading that bonding ethernet devices is supposed to be easy. I have followed several tutorials and tried a few things on my own, and I can't seem to get it right. I currently have a setup that allows me to ping internally on my network, but when I try to ping externally I get the error:
connect: Network is unreachable
Before I bonded the interface my connection worked fine. Here are my config files. I am running centos 5.3 on a virtual machine.
/etc/sysconfig/network-scripts/ifcfg-bond0
It's more just straight networking rather than Linux related - possibly a rather large gap in my knowledge.Been asked to set up a Linux firewall / router for a friend, and he wants an external address on his PC and his firewall. He's got a /30... ISP seem to think this is normal. My understanding was different router interfaces should be in different subnets when it comes to addressing. eg:
WAN 1.1.1.1LAN 192.168.1.1/24 (then PC 192.168.1.2)This is how I'd set up an enterprise router, and from memory how I had to regurgitate networking for my Cisco exams. But they would have been large devices within a BGP environment, how does this compare to home use? Can you have interfaces on the same router within in the same subnet? I'm looking for the real world answer - not my CCNA answer or the config addressing scheme I just blindly followed...Is this right or can different interfaces be in the same subnet?If they can be in different subnets How would you do that with an eternal 1.1.1.1/30? egWAN 1.1.1.1LAN (what IP?) then PC 1.1.1.2Would you actually just set the router up as a /29 and then use the additional IP addresses for the LAN NIC on the router? (yes I'm aware you couldn't route to where it had actually been allocated)
Then with that how do I route the 1.1.1.1/30? Which interface do I point it at? Even with a /29 I'd still have a routing issue right?Or are the ISP expecting you to do something with NAT / port forwarding. (and yes I could do this to get it working - but I want to understand it better) I'm asking the question here, because rather than just wanting a magic file to get things working, I'd actually rather understand the principles rather than just press a button and have it work.
My laptop was working fine on wireless till the userinterface changed and it defaulted to ethernet and now it won't let me go back on wireless How do i disable ethernet?
I have 2 NICs on my computer connected to 2 different networks. On 1 interface i wish to disable the return of ping and on the other interface i wish to still allow ping. Is this possible? Both interfaces need to be able to send and receive data. I know that you can disable ping in /etc/sysctl.conf but is there a way to set the ping value for a specific interface only?
I have an ubuntu server that I am using as a gateway router. I have 2 nics in the machine and am using iptables to run the NAT.I now have a need to provide several machines on the inside of the network with their own public IP. So I created a virtual interface on the the WAN card and attempted a 1:1 nat across it, but it's not working. The virtual interface isn't even responding to pings. If I ping the ip of the virtual interface from the outside, it doesn't work. If I run a tcpdump on the interface, I can see ICMP request packets but no replies. Watching the same dump while pinging the non-virtual interface IP results in both request and reply packets. What's stranger the inbound ICMP request packets on the virtual interface seem to be hit or miss. Sometimes when I ping it, I can see the request packets hitting the interface, sometimes I get nothing.
I am working on a project that needs to know that an interface is up and has stable connectivity so I am told by higher ups that I am going to use ping to solve this. So I was writing something much like the ping program to send out ping packets to [URL]. The issue comes in the fact that I need to do this for every interface that is currently up. So I mimic the code from iputils version of ping to tie the socket to an interface and everything works well except I don't get any packets returned to me.
So digging further I opened wireshark and indeed my interface is getting the ICMP echo response packets but linux (ubuntu 8.10) isn't passing the packet up. So I tried running "sudo ping -I ppp0 www.google.com" and the exact same behavior happens here too. (also seen but not solved here: [URL]. Also my iptables in empty and I also have a wifi connection which is my default gateway in addition to the 3G ppp0 connection. It is not a routing issues cause when I collect packet on the given interface I see the packets that should be there.
The entry in the routing table for ppp0 is: (dst)68.28.xx.yy (gateway)0.0.0.0 (genmask)255.255.255.255 (flags)UH (metric)0 (ref)0 (use)0 (Iface)ppp0
So I was wondering if anyone would either A) be willing to spend a minute to check and see if they are able to ping with the -I option and if so let me know if their routing has entries for that interface (route -n) or B) if anyone has any ideas on why the system isn't passing the packet up to the socket?
I am unable to connect to the internet at work through the ethernet, previously working. I am not sure of the origin of this problem -- the result of an update or something that I did to the network settings.
In the NetworkManager applet is says Auto Ethernet is connected but I can't ping anything by name or ip address.
The wireless still works without any problem.
I have tried following some of the other threads on similar issues but the fixes proposed on them have not helped to date. And I may have mangled some of the settings files somewhat.
I've a server with two NIC. Sometimes the server becomes unreacheable. When this happens, I did a reboot and then it works.. but when looking at the interface status (with ifconfig command), I've notice several "errors".So, I've tried to change NIC.. but the problem occurs again.I've also changed the port on the switch.
I'm just installed Fedora 14 64-bits into a server which come with Multiple network interfaces, I'm found that the naming of each network interface is not in sequences in what I'm thought (e.g: the on board network interfaces name as Eth5 and Eth6, the additional card ethernet port name from Eth0 - Eth1 and etc). How to name the interfaces as what I wish to? What should I install to allow me to rename the interfaces.
I tried to setup a connection as below but face some problem (probably route table setting incorrect). Hope if anyone could shed the light.Basically, I have a PC1 (.1) and PC2 (.130) connected via a PC router. PC router has 2 interfaces: wireless (.2) and ethernet (.129). PC1 connects to PC router via LAN cable/wireless interface, while PC2 connects to PC router via cable. This settings try to simulate 1 wireless connection along the path (and because PC2 is too old to support wireless interface, we need a PC router). These interfaces are all under same 172.16.130.x subnet.
I'm trying to create EoIP interface on ubuntu so i can create a simple tunnel to my mikrotik router. Is there anyone know how or even done that? If EoIP is not possible, is there any other simple way?
I had already read and thought about doing it with OpenVPN, but when I read the community documentation for OpenVPN on ubuntu 10.10, I fear it won't connect the tunnel to mikrotik OpenVPN server, since OpenVPN on ubuntu uses 2 certificate and 2 key files(as i read on the docs), but mikrotik configuration, i can see only 1 certificate can be applied on. This confuses me and make me decide to use EoIP(but i can't find any tutorial/docs about it). I don't actually need the encryption and security, i just need to create a tunnel for ubuntu and mikrotik.
I'm running Ubuntu 11.04 on my desktop and the 70-persistent-net.rules file seems to be ignored. The ethernet interfaces are randomly shuffled around on boot (i have 4 interfaces). The original one was configured to have a static IP address (eth0). Is there anything that prevents the /etc/udev/rules.d/70-persistent-net.rules from being applied/used?
I have installed Fedora 14 about 4 days ago. I have set all the networking parameters like dns path, hostname etc. I have also ensured that I select the check-box for "Enable automatically when system starts" for eth0 interface. But weirdly, whenever I boot my system(or restart), I see that the interface eth0 is disabled. I have to manually enter as root and enable it, each time I boot/reboot my system. Why is this happening? Could you please suggest a way so that I have eth0 interface enabled always when I boot up?Also, my domainname also is not boot-persistent. What steps should I take to ensure that the domainname set once persists across reboots?
Currently I have a server which runs under centOS 5.6. It is dedicated to the VoIP application of my customer.I have a problem for which I have the solution but I didn't managed to achieve it.So, let me explain you the context.Here is the networking aspects of my environment
VoIP Provider_____Gateway_____________My server ADSL Provider____(non pingable) x.x.x.2 <====> A.A.A.1 <======> A.A.A.3
On FC10, when system boots up, it would auto detect (from DHCP) and configure eth0.
This does not happen on FC11, but I can manually configure the interface. The entries in /etc/sysconfig/network-scripts/ifcfg-eth0 are identical on both OS releases.
Don't know if this is related, but when I run System->Administration->Bootloader, it fails system-config-boot - missing module kudzu.
I have tried the command on ubantu workstation but it giving error : # mii-tools SIOCGMIIPHY on 'eth1' failed: operationsnot support. no MII interface found.
Internet not working. tell me anonymous of kudzu command in fedora.
eth0 is connecting via a Cisco switch to the outer world and eth1 and eth2 to internal networks.
eth1 and eth2 are working without problems all the time, however I have the following incident with eth0: although it is RUNNING, and after it works for some time (like some hours), it then becomes incapable of sending/receiving traffic. Here are more details, after the interface exits the RUNNING state:
I am wondering whether there this is because there is not traffic for some time and so the interface goes to idle. Is there any parameter I can tweak in RHEL6 to force the interface to remain RUNNING even in the absence of traffic (although I would be expecting that some 802.1 packets would be periodically flowing from the switch to the interface)?
Im new to this Cent OS as well as Unix based Operating Systems. Please help me with the below requirement.My Company is a Serach based data collector from different websites like Google, Yahoo & many more..Collecting the data from the search engines is a issue for right now.. As many search engines are blocking us to search and retrive the date after some time.. Taking Google as example : Google Blocks my IP after every 30 mins of retriving.. Application is a Windows based & using Windows 2003 as the Server.. Would like to give a Dynamic IP to the Windows Server via 1st Ethernet Card & enable the 2nd Ethernet for the Internet..Eg : eth 0 = 192.168.1.1eth 1 = x.x.x.x (Public IP)I have check with few of my friends who told me to get Linux or Cent OS installed to do a interface disable after given time & enable a secondary device or virtual interface the next couple of seconds... Like eg : eth 1 = x.x.x.1 eth 2 = x.x.x.2 eth 3 = x.x.x.3 so on.. Would like the eth 1 to be down after 60 mins & eth 2 o be up & again after 60 mins of up time eth 2 down & eth 3 up.
I am using an virtual machine. where I need to ping from one machine to another. earlier I was able to ping. But after going to google.com once, I cannot ping back to this machine.
But if I gave ping -I eth1 <IP> then I can ping.
I cannot install any package, so tell me solution which includes not installing any package.
I'm having trouble getting Fedora 15 to work with my em/bond/br configuration as it did in Fedora 14. I've got a bonding interface over em1 and em2, and then individual bond vlans (bond0.2, bond0.10, etc...), I then have a similarly named bridge interface (br0.2, br0.10, etc...) for kvm. Regardless of if I assign bond0.2 or br0.2 the IP address, I am unable to ping the gateway, but the native untagged VLAN (bond0, br0) has no issue at all.
To try to isolate this issue down to the switch or Fedora 15, I went with a plain em setup and configured em1, and em1.2 which works as expected. This issue has been plaguing me since Fedora 15 came out and I'm about to trash it and go back to Fedora 14 (Which didn't have this issue), but some of the new kvm features have kept me trying to get over this incomprehensible hump.
I have a rack of IBM x335s. They are all running 802.1Q over 802.3ad, that is VLAN trunking over two physical Ethernet bonded with LACP. Its been working fine on all of the servers for years.One server recently had its hard drives upgraded and CentOS reinstalled. It is a minimal install. The server has been reconfigured exactly as it was and has been verified and compared against the other working servers.It will not enslave the interfaces on boot. You can manually enslave and then bring up the network, so that much of the configuration is verified. It simply refuses to enslave... which fails everything that comes after thatHere is a snippet of the message log on the machine that does not enslave on boot
Jan 31 22:28:57 vm3srvp01 kernel: NET: Registered protocol family 10 Jan 31 22:28:57 vm3srvp01 kernel: lo: Disabled Privacy Extensions Jan 31 22:28:57 vm3srvp01 kernel: IPv6 over IPv4 tunneling driver
I am not sure if this is a problem with my router or a problem with the fedora 12 server i am using.
- internal here means within my home network, external means from outside it.
I can access my server and webpage:
internally using http://192.168.0.192 externally using http://MYDNS___EXAMPLE.com
but I cannot access the server internally using http://MYDNS___EXAMPLE.com
This is a pain in the butt because I have all of my client software setup with an internal and external access point (ssh, proFTPd, http, MS Access database connected to MySql, etc.)
I booted an OS from live disk. The OS boots successfully but eth0 fails to start. I checked lspci , it outputs the exact ethernet card details :(Broadcom Netextreme II BCM5716 Gigabyte ethernet)
Ifconfig -a returns only loop back interface lo and sit0. The output of dmesg | grep -i 'eth' is :"Netfront : initializing network ethernet driver"
I was also able to find the kernel module bnx2.ko. I did insmod and lsmod lists it. I am not sure what exactly is the issue or how to debug this problem. Any assistance or pointer will be helpful. Thanks for your time.
