[Lustre-discuss] A very strange problem with lustre

高兴 zhi.xing.gao at hotmail.com
Mon Nov 12 07:04:30 PST 2007


Hi,I met a very strange problem with lustre-1.6.3(the rhel-2.6-x86_64 version),when I installed lustre in the cluter nodes, the nodes which installed lustre cann't ping each other, the ping command always reply "Destination Host Unreachable", each of the nodes has three netcard(two are 1000M Intel(R) PRO,the other is 100M Intel(R) PRO,the first 1000M netcard's ip address is configured as 192.168.0.*,the other is used as external netcard,and the 100M netcard is configured as 192.168.1.*). 
If the nodes boot with the stardard kernel, the network has no problem. 
It seems that the nodes cann't fetch mac address of the other.
I am confused by the problem,what is the matter at all? 

Let me show you the problem with a example:
nodeA (with the lustre kernel):eth0:192.168.1.2;eth1:192.168.0.2;eth2:202.117.10.*
nodeB (with the lustre kernel):eth0:192.168.1.3;eth1:192.168.0.3;eth2:202.117.10.*

nodeA:#ping 192.168.0.3    the ping command failed
nodeA:#ping 192.168.1.3    the ping command ok
nodeA:#ping 202.117.10.*   the ping command ok

nodeB:#ping 192.168.0.2    the ping command failed
nodeB:#ping 192.168.1.2    the ping command ok
nodeB:#ping 202.117.10.*   the ping command ok

the lustre packets I installed were listed as follow:
kernel-lustre-smp-2.6.9-55.0.9.EL_lustre.1.6.3.x86_64.rpm
lustre-1.6.3-2.6.9_55.0.9.EL_lustre.1.6.3smp.x86_64.rpm
lustre-debuginfo-1.6.3-2.6.9_55.0.9.EL_lustre.1.6.3smp.x86_64.rpm
lustre-ldiskfs-3.0.2-2.6.9_55.0.9.EL_lustre.1.6.3smp.x86_64.rpm
lustre-ldiskfs-debuginfo-3.0.2-2.6.9_55.0.9.EL_lustre.1.6.3smp.x86_64.rpm
lustre-modules-1.6.3-2.6.9_55.0.9.EL_lustre.1.6.3smp.x86_64.rpm

_________________________________________________________________
手机也能上 MSN 聊天了,快来试试吧!
http://mobile.msn.com.cn/




More information about the lustre-discuss mailing list