OpenVZ Forum: Support » ARP? Short network outages on OpenVZ HOST when starting or stopping OVZ nodes (netw bridging)

Home » General » Support » ARP? Short network outages on OpenVZ HOST when starting or stopping OVZ nodes (netw bridging)

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

ARP? Short network outages on OpenVZ HOST when starting or stopping OVZ nodes (netw bridging) [message #27409]

Sun, 17 February 2008 16:38

vazir
Messages: 12
Registered: October 2007

Junior Member

Does anyone seen that glitch, when the virtual host started or stopped, in 95 cases (with me - the main host looses network (seems ARP tables getting emptied) - and restores in ~30 seconds. None of virtual nodes oin this host looses connectivity - just hardware node. For networking I use bridged networking (veth)
The strange thing is that this happens not on the every node.
Starting/stopping some of the nodes does not break networking os the host at all.

[Updated on: Sun, 17 February 2008 18:20]

Report message to a moderator

Re: ARP? Short network outages on OpenVZ HOST when starting or stopping OVZ nodes (netw bridging) [message #27456 is a reply to message #27409]

Mon, 18 February 2008 08:48

den
Messages: 494
Registered: December 2005

Senior Member

I think that your see a MAC address change on your bridge. In this case a DHCP will re-negotiate the HWN IP.

Pls check this. eth0 MAC should be the smallest beyond ones inserting to a bridge. If this does not help, pls explain you setup in more details including exact kernel version you are using.

Regards,
Den

Report message to a moderator

Re: ARP? Short network outages on OpenVZ HOST when starting or stopping OVZ nodes (netw bridging) [message #27458 is a reply to message #27456]

Mon, 18 February 2008 09:07

vazir
Messages: 12
Registered: October 2007

Junior Member

Hi Den!

I don't use DHCP in my network, a little confused how DHCP is involved in this.

My setup is as follows:

Behaves the same on:
Linux vz1 2.6.18-fza-028stab051.1-686 #1 SMP Fri Dec 7 17:44:02 CET 2007 i686 GNU/Linux (P4 netburst)

Linux vz2 2.6.22-ovz005.1 #1 SMP Thu Nov 15 15:35:00 MSK 2007 x86_64 GNU/Linux (Core2 quad)

br0 Link encap:Ethernet HWaddr 00:18:51:52:7C:8C
eth1 Link encap:Ethernet HWaddr 00:50:04:04:0A:D7
veth116.0 Link encap:Ethernet HWaddr 00:18:51:99:F3:C7
veth122.0 Link encap:Ethernet HWaddr 00:18:51:89:DB:CF
veth124.0 Link encap:Ethernet HWaddr 00:18:51:8B:9C:ED
veth126.0 Link encap:Ethernet HWaddr 00:18:51:D0:EB:B7
veth130.0 Link encap:Ethernet HWaddr 00:18:51:A2:C8:5F

Do you think having eth1 HWaddr 00:50:04:04:0A:D7 - might be a problem?

config of one of the VE, which affects HN

ONBOOT="yes"

# UBC parameters (in form of barrier:limit)
# Primary parameters
AVNUMPROC="40:40"
NUMPROC="512:512"
NUMTCPSOCK="80:80"
NUMOTHERSOCK="80:80"
VMGUARPAGES="6144:2147483647"
# Secondary parameters
KMEMSIZE="268435456:268435456"
TCPSNDBUF="1924288:1924288"
TCPRCVBUF="1924288:1924288"
OTHERSOCKBUF="132096:336896"
DGRAMRCVBUF="132096:132096"
OOMGUARPAGES="6144:2147483647"
# Auxiliary parameters
LOCKEDPAGES="32:32"
SHMPAGES="8192:8192"
PRIVVMPAGES="262144:262144"
NUMFILE="2048:2048"
NUMFLOCK="100:110"
NUMPTY="16:16"
NUMSIGINFO="256:256"
DCACHESIZE="1048576:1097728"

PHYSPAGES="0:2147483647"
NUMIPTENT="128:128"

# Disk quota parameters (in form of softlimit:hardlimit)
DISKSPACE="10485760:10485760"
DISKINODES="2000000:2000000"
QUOTATIME="0"

# CPU fair sheduler parameter
CPUUNITS="1000"
VE_ROOT="/home/vzmount/forum2"
VE_PRIVATE="/home/vzhosts/forum2"
OSTEMPLATE="debian_forum_backup"
ORIGIN_SAMPLE="template.10GB"
NAME="forum2"
HOSTNAME="forum2"
NETIF=" ifname=eth0,mac=00:18:51:1A:6E:66,host_ifname=veth130.0,host _mac=00:18:51:A2:C8:5F "
CUSTOM_CONFIG="yes"

Report message to a moderator