VIF not reachable after takeover

[FRANKLUNA]

I have a FAS 3240 being initialized. I have created two vifs on head one and two using the onboard e0a and e0b on each controller. I have configured two vlans on each controller vif and set them up to be partners to each other. The problem I am having is that when I do a manual takeover, the vlan interfaces from the failed controller are not reachable after the takeover. I can ping the partner addresses on the controller taking over but I cant from any clients on the network. Can anyone explain this behavior? This is an initial setup.

hostname host1

ifgrp create lacp vif1FCoE -b ip e1b e1a

ifgrp create lacp vifh1cifs -b ip e0b e0a

vlan create vifh1cifs 32 34

ifconfig vif1FCoE `hostname`-vif1FCoE netmask 255.255.255.0 partner vif2FCoE mtusize 1500 trusted wins up

ifconfig e0M `hostname`-e0M netmask 255.255.255.0 mtusize 1500 trusted wins mediatype 100tx-fd flowcontrol full up

ifconfig vifh1cifs-34 `hostname`-vifh1cifs-34 netmask 255.255.255.0 partner vifh2cifs-34 mtusize 1500 trusted wins up

ifconfig vifh1cifs-32 `hostname`-vifh1cifs-32 netmask 255.255.255.0 partner vifh2cifs-32 mtusize 1500 trusted wins up

route add default 10.237.34.1 1

routed on

options dns.enable off

options nis.enable off

savecore

hostname host2

ifgrp create lacp vif2FCoE -b ip e1a e1b

ifgrp create lacp vifh2cifs -b ip e0a e0b

vlan create vifh2cifs 32 34

ifconfig vif2FCoE `hostname`-vif2FCoE netmask 255.255.255.0 partner vif1FCoE mtusize 1500 trusted wins up

ifconfig e0M `hostname`-e0M netmask 255.255.255.0 mtusize 1500 trusted wins mediatype 100tx-fd flowcontrol full up

ifconfig vifh2cifs-32 `hostname`-vifh2cifs-32 netmask 255.255.255.0 partner vifh1cifs-32 mtusize 1500 trusted wins up

ifconfig vifh2cifs-34 `hostname`-vifh2cifs-34 netmask 255.255.255.0 partner vifh1cifs-34 mtusize 1500 trusted wins up

route add default 10.237.34 1

routed on

options dns.enable off

options nis.enable off

savecore

[scottgelb]

Sounds like a switch configuration. Ontap is failing the interfaces over. Is the switch config not allowing 2 ips or macs from the same port?

[FRANKLUNA]

I cant imagine this is the case since each controller is connected using 2 uplink ports in a LACP trunk on e0a and e0b and it passes traffic (two vlans) until it is failed over. Im really not sure what to think of this. Ive had 4 Netapp level one guys tell me my RC config looks fine. Maybe this a default route issue. I have e0M on a seperate subnet as the e0A and e0B interfaces. Any thoughts?

[scottgelb]

Routing is possible. But should be the same on failover. Does "partner netstat -rn" how the routes? Does "partner ping gatewayIP" work? You can trace route too in ontap.

Sent from my iPhone 4S

[scottgelb]

Rc looks good to me too. I did run into an issue at a customer where they didn't allow more than one mac address per port and that restriction caused an issue. May not be the case here but worth checking.

[FRANKLUNA]

Ill have to give this a try tomorrow morning, its just frustrating when nothing seems out of place. We have used two switches so far and they both have had the same problem. The default route is the network that one of the vlan interfaces is on so Im not sure why during failover at least one would not work. Im stumped at this point.

[VKALVEMULA]

can you check this ...     options dns.enable off

can you change this to    options dns.enable on

and check.

[FRANKLUNA]

I have the 3240 on a bench isolated, I am cabled in. I have also enabled dns before but it didn't help.

[VKALVEMULA]

check under your /etc/hosts file

does both the filers have VIF IP address details

IP add(x.x.x.x)    VIF1

IP add(x.x.x.x)    VIF2

when filer reboots or takeover make sure you have all the VIF details in /etc/hosts file so that it can read and enable the appropiate VIFs.

i had the same issue after reboot as yours.. so

once i added the ip add and vif details under /etc/hosts.. all the VIFs were automatically onlien and enable after my 2nd test takeover and giveback.