Thanks Mike for getting back to me. As I mentioned earlier if eth2 is down we can still ping the array through other interfaces like
ping -I eth3 192.168.42.190 If I bring down the eth2 via "ifconfig eth2 down" then no issue and routing table gets updated itself but I am disabling the eth2 by disabling the port on switch so system routing doesnt change and If I change the routing metric of eth2 then it uses other interfaces but I dont think it is appropriate solution. My understanind is discovery should do the job when we define the iface parmater in the command line but follwoing command still try to connect array via eth2 which is the first and default interface to connect array. iscsiadm -m discovery -t st -p 192.168.42.190:3260 -I eth6 Should I try the git code Farhan On Jul 23, 12:12 pm, Mike Christie <micha...@cs.wisc.edu> wrote: > On 07/21/2011 11:28 PM, Farhan Ahmed wrote: > > > > > > > Hi Forum, > > > I am having very weird open-iscsi issue. Could you guys help me to > > find out the reason? > > > There are 4 network interfaces(eth2/eth3/eth5/eth6) for iSCSI traffic, > > eth2/eth5 are patched into one HP switch and eth3/eth6 are patched to > > second switch. In normal scenario all seem working fine. > > We are testing all scenarios before getting the box into prod. I am > > testing scenario in case switch fails but somehow iscsiadm cant > > perform the discovery. > > > All interfaces are on same subnet and eth2 becomes the default for > > that subnet. > > > If I disable eth2 from the HP switch I cant ping the array but I can > > ping the array via eth3/eth5/eth6 but I cant perform iscsiadm > > discovery through interfaces eth3/eth5/eth6 > > The current version of open-iscsi uses the default routing for discovery > so if you cannot even do a normal ping like below then iscsiadm > discovery is not going to work. The code in git is able to use the iface > binding info to bind the discovery session to a interface. > > When you disable eth2 shouldn't the network just figure out that the > other interface eth5 is on the same subnet and drop down to that? Do you > have to update the routing tables by hand then? > > > > > > > disable eth2 on switch > > > [root@nfs02 ~]# ping 192.168.42.190 > > PING 192.168.42.190 (192.168.42.190) 56(84) bytes of data. > > From 192.168.42.70 icmp_seq=2 Destination Host Unreachable > > From 192.168.42.70 icmp_seq=3 Destination Host Unreachable > > From 192.168.42.70 icmp_seq=4 Destination Host Unreachable > > > [root@nfs02 ~]# ping -I eth3 192.168.42.190 > > PING 192.168.42.190 (192.168.42.190) from 192.168.42.71 eth3: 56(84) > > bytes of data. > > 64 bytes from 192.168.42.190: icmp_seq=1 ttl=255 time=0.166 ms- Hide quoted > > text - > > - Show quoted text -- Hide quoted text - > > - Show quoted text - -- You received this message because you are subscribed to the Google Groups "open-iscsi" group. To post to this group, send email to open-iscsi@googlegroups.com. To unsubscribe from this group, send email to open-iscsi+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/open-iscsi?hl=en.
