OpenVZ Forum


Home » General » Support » [solved] arpsend issue - URGENT
[solved] arpsend issue - URGENT [message #32497] Tue, 12 August 2008 20:17 Go to next message
Skylar is currently offline  Skylar
Messages: 5
Registered: August 2008
Junior Member
Quote:

[root@lithium ~]# vzctl restart 1520
Restarting VE
Stopping VE ...
VE was stopped
VE is unmounted
Starting VE ...
VE is mounted
Adding IP address(es): 69.162.73.103 69.162.73.105 69.162.73.106 69.162.73.107 69.162.73.108 69.162.73.104
arpsend: 69.162.73.103 is detected on another computer : 00:14:1c:1d:67:80
vps-net_add WARNING: arpsend -c 1 -w 1 -D -e 69.162.73.103 eth1 FAILED
arpsend: 69.162.73.105 is detected on another computer : 00:14:1c:1d:67:80
vps-net_add WARNING: arpsend -c 1 -w 1 -D -e 69.162.73.105 eth1 FAILED
arpsend: 69.162.73.106 is detected on another computer : 00:14:1c:1d:67:80
vps-net_add WARNING: arpsend -c 1 -w 1 -D -e 69.162.73.106 eth1 FAILED
arpsend: 69.162.73.107 is detected on another computer : 00:14:1c:1d:67:80
vps-net_add WARNING: arpsend -c 1 -w 1 -D -e 69.162.73.107 eth1 FAILED
arpsend: 69.162.73.108 is detected on another computer : 00:14:1c:1d:67:80
vps-net_add WARNING: arpsend -c 1 -w 1 -D -e 69.162.73.108 eth1 FAILED
arpsend: 69.162.73.104 is detected on another computer : 00:14:1c:1d:67:80
vps-net_add WARNING: arpsend -c 1 -w 1 -D -e 69.162.73.104 eth1 FAILED
Setting CPU limit: 65
Setting CPU units: 1000
Configure meminfo: 393216
Set hostname: server.xlhost.ws
File resolv.conf was modified
Setting quota ugidlimit: 1000
VE start in progress...


After I do this and try to get into the system, I no longer can as it gives me a fork: out of memory error, and HyperVM shows that the server is in an offline without-memory state.

I had the DC change eth0 and eth1 around, as I was getting an "arpsend WARNING" issue instead of the new "vps-net_add WARNING"

This only happens on my newest machine; all of the other 3 I can make a VPS with the exact same IPs and they work 100% fine; I can reboot without seeing this error

I need this fixed ASAP


Edit: Fairly new system, Centos 5.2 64bit; not sure on the OpenVZ version though, no clue how to check >.<

Edit again: If I remove the IPs from the specified VPS, reboot the VPS; THEN add the IPs then the server works just fine and the IPs work; the VPS only won't work when I try to reboot it.

[Updated on: Tue, 12 August 2008 21:22] by Moderator

Report message to a moderator

Re: arpsend issue - URGENT [message #32498 is a reply to message #32497] Tue, 12 August 2008 20:25 Go to previous messageGo to next message
khorenko is currently offline  khorenko
Messages: 533
Registered: January 2006
Location: Moscow, Russia
Senior Member
Hello.

Quote:

I no longer can as it gives me a fork: out of memory error

If you get out of memory failures - check failcounters in /proc/user_beancounters file and resolve them if any.

Quote:

arpsend: 69.162.73.104 is detected on another computer : 00:14:1c:1d:67:80
vps-net_add WARNING: arpsend -c 1 -w 1 -D -e 69.162.73.104 eth1 FAILED

Can you just ping the address in question while the VE is down? If it's pingable - you'll have to resolve the IP duplication.
Just check the node with MAC 00:14:1c:1d:67:80 - arpsend reported that the address is used on it (probably on some it's VE).

--
Konstantin


If your problem is solved - please, report it!
It's even more important than reporting the problem itself...
Re: arpsend issue - URGENT [message #32500 is a reply to message #32498] Tue, 12 August 2008 20:27 Go to previous messageGo to next message
Skylar is currently offline  Skylar
Messages: 5
Registered: August 2008
Junior Member
We have already checked and the IPs are not in use. I edited my post above to explain how I can get it to work; check that then see if you know what may have caused this.

Quote:

Edit again: If I remove the IPs from the specified VPS, reboot the VPS; THEN add the IPs then the server works just fine and the IPs work; the VPS only won't work when I try to reboot it.

[Updated on: Tue, 12 August 2008 20:31]

Report message to a moderator

Re: arpsend issue - URGENT [message #32502 is a reply to message #32500] Tue, 12 August 2008 20:46 Go to previous messageGo to next message
khorenko is currently offline  khorenko
Messages: 533
Registered: January 2006
Location: Moscow, Russia
Senior Member
And what is the node with MAC 00:14:1c:1d:67:80?
Is that the same node where you are trying to restart VE or another node?


If your problem is solved - please, report it!
It's even more important than reporting the problem itself...
Re: arpsend issue - URGENT [message #32503 is a reply to message #32502] Tue, 12 August 2008 20:48 Go to previous messageGo to next message
Skylar is currently offline  Skylar
Messages: 5
Registered: August 2008
Junior Member
The DC won't tell me which it is so I do not know.


Now I think that this newer node is on a different router, and I have IP pooling so I think it is trying to go through the wrong router to the VPS...

The weird part is when I go to stop the VZ it takes like 3-5 minutes just to stop.
Re: arpsend issue - URGENT [message #32505 is a reply to message #32503] Tue, 12 August 2008 21:03 Go to previous messageGo to next message
Skylar is currently offline  Skylar
Messages: 5
Registered: August 2008
Junior Member
Is there any way to re-install OpenVZ without having to move the 11 VPS clients off of the node or without loosing data? I think the best bet may be to just re-install
Re: arpsend issue - URGENT [message #32506 is a reply to message #32503] Tue, 12 August 2008 21:07 Go to previous messageGo to next message
khorenko is currently offline  khorenko
Messages: 533
Registered: January 2006
Location: Moscow, Russia
Senior Member
1) you can check if that MAC belongs to YOUR node and if not - then another physical node has this MAC.
2) you can check your arp table - probably it'll have needed entry.


If your problem is solved - please, report it!
It's even more important than reporting the problem itself...
Re: arpsend issue - URGENT [message #32508 is a reply to message #32506] Tue, 12 August 2008 21:09 Go to previous messageGo to next message
Skylar is currently offline  Skylar
Messages: 5
Registered: August 2008
Junior Member
<Myself> I think one of the files may have been corrupt wile installing but the installer didnt catch it
<lx_lxsales> in that case, do rpm -e vzctl
<lx_lxsales> yum -y install vzctl


I ran the re-install of vzctl and it worked :D:D:D:D:D:D
Re: arpsend issue - URGENT [message #32510 is a reply to message #32508] Tue, 12 August 2008 21:22 Go to previous message
khorenko is currently offline  khorenko
Messages: 533
Registered: January 2006
Location: Moscow, Russia
Senior Member
Well, my congratulations!
At the same time such a 'corruption' seems strange to me. Just in case - next time you can check if all files from rpm are ok/not corrupted by 'rpm -V <package>'.

--
Konstantin


If your problem is solved - please, report it!
It's even more important than reporting the problem itself...
Previous Topic: reboot inside vps
Next Topic: VE Fails to start with dmesg error "err=-11". 64bit el5 patch.
Goto Forum:
  


Current Time: Wed Feb 28 10:30:20 GMT 2024

Total time taken to generate the page: 0.02521 seconds