Backup data lan card

[Christopher McCray_1]

I am configuring 2 N-4000 servers for a cluster. Each server has 4 lan cards configured as follows:
Lan0 = primary heartbeat (10/100)
Lan1 = unconfigured, backup data (10/100/1000)
Lan2 = secondary heartbeat (10/100/1000)
Lan3 = primary data (10/100/1000)

doing a cmquerycl indicates lan1 as a possible backup for lan3, which is our intention, and we are under the assumtion that, in the case of lan3 experiencing failure, MCSG will configure lan1 with its information. Is this assumtion correct? If so, is there some configuration parameter we missed to enable this because we have tried to test this by pulling lan3's cable out, but nothing happens. Thanks in advance for your help.

Chris

[melvyn burnard]

provided the configuration is correct, then yes this should happen. Might be a silly question, but have you configured a package that monitors that subnet yet?
you could also do a cmscancl and post the output file here.

[Sridhar Bhaskarla]

You may need to check the following

1) Check your package configuration file to see if you have this

NET_SWITCHING_ENABLED YES


2) Verify the SUBNET on the primary LAN is being monitored by verifying the package configuration file

SUBNET YOUR_PRIMARY_SUBNET

If you have these two in place, this will work.

-Sri

[Christopher McCray_1]

Melvyn,
Sorry it took so long, but here is the output of the cmscancl. Unfortunately, I had to "sanitize" the IP addresses, but it should still be understandable. Thank you again for your help.

Sridhar,

I have the LOCAL_LAN_FAILOVER_ALLOWED set to YES and SUBNET is correct. Thanks

[melvyn burnard · Accepted solution]

according to th ebinary file info in the cmscancl output, you do not have a backup lan!

Network ID 2:
ppa: 3
old_ppa: 0
mac addr: 0x00306e038ea7
hardware path: 0/4/0/0
network interface name: lan3
subnet: 10.12.13.0
subnet mask: 255.255.255.0
ip address: 10.12.13.88
flags: 1 (Heartbeat Network)
bridged net ID: 1

Network ID 1:
ppa: 0
old_ppa: 0
mac addr: 0x00306e035982
hardware path: 0/0/0/0
network interface name: lan0
subnet: 13.14.15.16
subnet mask: 255.255.255.252
ip address: 13.14.15.17
flags: 1 (Heartbeat Network)
bridged net ID: 2

Network ID 3:
ppa: 2
old_ppa: 0
mac addr: 0x00306e0e17d3
hardware path: 1/8/0/0
network interface name: lan2
subnet: 13.14.15.32
subnet mask: 255.255.255.240
ip address: 13.14.15.33
flags: 1 (Heartbeat Network)
bridged net ID: 3

Node ID 2:
Node name: dmc31
first lock pv name: /dev/dsk/c5t0d2
first lock disk interface type: c720

Network ID 1:
ppa: 3
old_ppa: 0
mac addr: 0x00306e039e3e
hardware path: 0/4/0/0
network interface name: lan3
subnet: 10.12.13.0
subnet mask: 255.255.255.0
ip address: 10.12.13.92
flags: 1 (Heartbeat Network)
bridged net ID: 1

Network ID 2:
ppa: 0
old_ppa: 0
mac addr: 0x001083963135
hardware path: 0/0/0/0
network interface name: lan0
subnet: 13.14.15.16
subnet mask: 255.255.255.252
ip address: 13.14.15.18
flags: 1 (Heartbeat Network)
bridged net ID: 2

Network ID 3:
ppa: 2
old_ppa: 0
mac addr: 0x00306e0e1701
hardware path: 1/8/0/0
network interface name: lan2
subnet: 13.14.15.32
subnet mask: 255.255.255.240
ip address: 13.14.15.34
flags: 1 (Heartbeat Network)
bridged net ID: 3



the lan you expect to have as standby i snot there, and is probaly the result of the binary being applied before this lan was actually there.
you need to reapply the binary with this info in the ascii file and the links all ok.

[Christopher McCray_1]

Melvyn, you are the MAN!!!!! As soon as we got your final response, we looked in our cluster.conf file and, sure enough,
NETWORK_INTERFACE lan1
was not there. We added it, and rock and roll!! I apologize for not getting back earlier (much problems elsewhere), but the ten points you are getting from me is not near enough. Thanks to you to Sridhar also.

Good luck to you all!
Chris