OpenVZ Forum: Support » eth1 on VPS

Home » General » Support » eth1 on VPS

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

eth1 on VPS [message #7397]

Fri, 13 October 2006 00:27

ferp2
Messages: 29
Registered: May 2006

Junior Member

Hi,

I have the following openvz setup:

2.6.8-022stab064-up on Debian Sarge.

HN and VPS (101, 102, 103, 104) on subnet 192.168.0.0

I decided to add Samba to VPS 104. Reading the forums, I learned that a VPS cannot send and receive broadcast signals, needed by Samba. Since I had 2 nic cards, I did:

Quote:

vzctl set 104 --netdev_add eth1 --save

Logged into VPS 104, I put the following in /etc/network/interfaces.template:

Quote:

auto eth1

iface eth1 inet static
address 192.168.0.104
netmask 255.255.255.0
network 192.168.0.0
broadcast 192.168.0.255
gateway 192.168.0.254

Logged into the HN, I removed any config for eth1.

2 Problems:

1. Logged into VPS 104, I cannot ping the HN or the other, regular VPSs. Logged into the HN, I likewise cannot ping VPS 104. How do I enable communication between VPS 104 and the HN/other VPSs.

2. Logged into VPS 104, if I do ifconfig -a, I see that both eth1 and venet0 are up. Is this right? Shouldn't venet0 be turned off when running VPS 104 from eth1. If this is true how do I prevent venet0 from getting configured when restarting VPS 104.

Thank you.

Report message to a moderator

Re: eth1 on VPS [message #7403 is a reply to message #7397]

Fri, 13 October 2006 05:39

Vasily Tarasov
Messages: 1345
Registered: January 2006

Senior Member

To get rid of venet device look at these thread http://forum.openvz.org/index.php?t=tree&th=270&mid= 1489&&rev=&reveal=

However the presense of venet device should not break network connectivity.

Quote:

Logged into VPS 104, I cannot ping the HN or the other, regular VPSs. Logged into the HN, I likewise cannot ping VPS 104.

Can you ping anyone else in network?
Please, do ifconfig -a and route -nv in VE and on HN and post the results here.

Thanks!

Report message to a moderator

Re: eth1 on VPS [message #7452 is a reply to message #7403]

Fri, 13 October 2006 15:06

ferp2
Messages: 29
Registered: May 2006

Junior Member

Hello Vass,

Thank you for your response.

To answer your requests:

Quote:

Can you ping anyone else in network?

Quote:

Please, do ifconfig -a and route -nv in VE and on HN and post the results here.

Please note that I have encounter new problems since rebooting the HN, which I will detail first before answering your requests:

Problem 1:
The booting sequence proceeded normally until booting VE 104. At this point, I received the following message.

Quote:

martian source 192.168.0.255 from 192.168.0.104,
on dev eth0
|| header: ff:ff:ff:ff:ff:ff:00:50:fc:72.56.44.08:00

Shortly after the HN system became unresponsive and required a cold reboot.

Upon rebooting a second time, the "martian source" message did not reappear, and the HN remained responsive.

Problem 2:
I could neither ssh nor ping VE 104. To remedy this, I did:

Quote:

$ cd /vz/private/104/etc/network
$ mv interfaces.template interfaces.template.off

I then rebooted VE 104 and was able to log in successfully.

Here are the details of ifconfig -a and route -nv:

ifconfig -a

Quote:

eth1 Link encap:Ethernet HWaddr 00:50:FC:72:56:44
BROADCAST MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:61 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:0 (0.0 b) TX bytes:10248 (10.0 KiB)
Interrupt:11 Base address:0xd400

lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:86 errors:0 dropped:0 overruns:0 frame:0
TX packets:86 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:11060 (10.8 KiB) TX bytes:11060 (10.8 KiB)

venet0 Link encap:UNSPEC HWaddr 00-00-FF-00-00-00-00-00-00-00-00-00-00-00-00-00
inet addr:127.0.0.1 P-t-P:127.0.0.1 Bcast:0.0.0.0 Mask:255.255.255.255
UP BROADCAST POINTOPOINT RUNNING NOARP MTU:1500 Metric:1
RX packets:187 errors:0 dropped:0 overruns:0 frame:0
TX packets:112 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:16380 (15.9 KiB) TX bytes:14481 (14.1 KiB)

venet0:0 Link encap:UNSPEC HWaddr 00-00-FF-FF-FF-FF-00-00-00-00-00-00-00-00-00-00
inet addr:192.168.0.104 P-t-P:192.168.0.104 Bcast:0.0.0.0 Mask:255.255.255.255
UP BROADCAST POINTOPOINT RUNNING NOARP MTU:1500 Metric:1

route -nv

Quote:

Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
191.255.255.1 0.0.0.0 255.255.255.255 UH 0 0 0 venet0
0.0.0.0 191.255.255.1 0.0.0.0 UG 0 0 0 venet0

Obviously, eth1 is not functioning in the present configuration. To make it functional, I will log into VE 104 and make the eth1 boot configuration file active again. Here are the details of that file:

/etc/network/interfaces.template

Quote:

auto eth1

iface eth1 inet static
address 192.168.0.104
netmask 255.255.255.0
network 192.168.0.0
broadcast 192.168.0.255
gateway 192.168.0.254

Now I will reboot VE 104 from the HN. Details are:

vzctl restart 104

Quote:

Restarting VPS
Stopping VPS ...
VPS was stopped
VPS is not mounted
Starting VPS ...
vzquota : (warning) Quota is running for id 104 already
VPS is mounted
Deleting IP address(es): 192.168.0.104
SIOCSIFFLAGS: Cannot assign requested address
Adding IP address(es): 192.168.0.104
Setting CPU units: 1000
Set hostname: vz4.lan.rmt
File resolv.conf was modified
VPS start in progress...

Note that I can no longer ping nor ssh into VE 104, which means I cannot reproduce the situation I was in when I started this thread.

Meanwhile, I have noticed 2 clues which may point to a solution to this problem. They are:

The martian source message I received before the HN became unresponsive and
SIOCSIFFLAGS: Cannot assign requested address

Here are a few details regarding my attempted network setup:

HN - 192.168.0.7 - eth0
VE 1 - 192.168.0.101 - eth0
VE 2 - 192.168.0.102 - eth0
VE 3 - 192.168.0.103 - eth0
=======================
VE 4 - 192.168.0.104 - eth1

Report message to a moderator

Re: eth1 on VPS [message #7459 is a reply to message #7397]

Fri, 13 October 2006 17:42

John Kelly
Messages: 97
Registered: May 2006
Location: Palmetto State

Member

Looks like eth1 and venet0 are both trying to grab the same ip address, .104. That would be a problem. Try changing eth1 to .204 and see if that helps. That may not be the final solution you are looking for, it's just a testing step to help narrow down the problem.

Report message to a moderator

Re: eth1 on VPS [message #7483 is a reply to message #7459]

Sun, 15 October 2006 16:28

ferp2
Messages: 29
Registered: May 2006

Junior Member

Thanks for your response, John. Your suggestion sounds like a good one. I have not yet had the chance to verify it, but would like to add the following about the VE 104 set up:

VE 104 was initially set up to use 192.168.0.4 on the venet0 interface. Later, I decided to add add eth1 to VE 104 using the

Quote:

vzctl set 104 --netdev_add eth1 --save

command.

If I log in to VE 104 and do ifconfig -a I can see the eth1 interface, although it is not functioning. Problems occur once I set eth1 within VE 104 to 192.168.0.4. Vass posted the following:

Quote:

However the presense of venet device should not break network connectivity.

This is true, perhaps, if venet0 has not been previously set to an ipaddress. In my case, I had already set the VE 104 ipaddress from the HN, as in:

Quote:

vzctl set 104 --ipadd 192.168.0.104 --save

Although I haven't yet determined if this is the solution ,I look forward to hearing what Vass has to say about this.

Ferp2

Report message to a moderator

Re: eth1 on VPS [message #7523 is a reply to message #7483]

Tue, 17 October 2006 01:51

ferp2
Messages: 29
Registered: May 2006

Junior Member

I did

Quote:

vzctl set 104 --ipdel 192.168.0.104 --save

and that seemed to solve some of the problems, however, there's still a conflict between the eth1 configuration and venet0, as I couldn't contact any hosts outside of the immediate subnet. The solution was to log in to VE 104 and do

Quote:

ifconfig venet0 down

How do I configure the VE 104 so that I don't have to manually shutdown venet0 each time I restart the server?

Thanks

Report message to a moderator

Re: eth1 on VPS [message #7531 is a reply to message #7523]

Tue, 17 October 2006 07:16

Vasily Tarasov
Messages: 1345
Registered: January 2006

Senior Member

Why don't you remove venet using http://forum.openvz.org/index.php?t=tree&goto=1516&# msg_1516,
how I offered in my previous answer? Rolling Eyes

HTH,
vass.

Report message to a moderator

Re: eth1 on VPS [message #7556 is a reply to message #7531]

Tue, 17 October 2006 13:04

ferp2
Messages: 29
Registered: May 2006

Junior Member

Hello,

In response to your recommendation:

Quote:

You need to comment out network setup in /etc/init.d/vz:

sed -i 's/$.*setup_net$$/#\1/' /etc/init.d/vz
/etc/init.d/vz restart

If you don't want restart service vz, then:
stop your VPSs,
ifconfig down venet0,
rmmod vznetdev

Correct me if I'm wrong, but it looks like the above solution removes the venet0 interface for all VEs, whereas I only want to remove venet0 from VE 4 (now using the eth1 interface), but leave the venet0 interface functioning on VEs 1, 2 and 3.

Thank you.

Report message to a moderator

Re: eth1 on VPS [message #7563 is a reply to message #7556]

Tue, 17 October 2006 14:08

John Kelly
Messages: 97
Registered: May 2006
Location: Palmetto State

Member

ferp2 wrote on Tue, 17 October 2006 09:04

I only want to remove venet0 from VE 4 (now using the eth1 interface), but leave the venet0 interface functioning on VEs 1, 2 and 3.

You can make your own custom VE scripts.

In /etc/vz/conf, make a new [xxx].conf file, edit it, and set OSTEMPLATE=mydist, where [mydist] is some arbitrary name you like.

Then, in /etc/vz/dists, copy one of the .conf files (whichever one suits your needs) and call it mydist.conf. Edit it, and set ADD_IP=mydist-add_ip.sh and DEL_IP=mydist-del_ip.sh.

Then in etc/vz/dists/scripts, choose a set of -add_ip.sh and -del_ip.sh scripts, and copy them to mydist-add_ip.sh and mydist-del_ip.sh, respectively.

Then start hacking the new scripts. This may require more effort than you had in mind, but once you learn how the scripts work, you can do most anything you want.

I don't know an easier way, maybe someone else does.

Report message to a moderator

Re: eth1 on VPS [message #7578 is a reply to message #7563]

Tue, 17 October 2006 18:58

ferp2
Messages: 29
Registered: May 2006

Junior Member

Sounds like the right way to go. I'll give it a try.

Thanks

Report message to a moderator

Previous Topic:	SOLVED Memory usage patch
Next Topic:	SOLVED always ipv6 ... :(

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sat Nov 02 10:36:49 GMT 2024

Total time taken to generate the page: 0.03237 seconds