Stanby LAN

[Vic S. Kelan]

Hi
Please does anyone have a tip I can use to verify which LAN is standby for another LAN card?

I have a SG package and noticed this:
NODE_NAME mainsrv
NETWORK_INTERFACE lan1
STATIONARY_IP 10.10.10.10
NETWORK_INTERFACE lan5
NETWORK_INTERFACE lan6
NETWORK_INTERFACE lan2
STATIONARY_IP 10.10.10.35
NETWORK_INTERFACE lan3
STATIONARY_IP 10.10.10.40
NETWORK_INTERFACE lan7
NETWORK_INTERFACE lan4
HEARTBEAT_IP 172.16.9.253
NETWORK_INTERFACE lan8
HEARTBEAT_IP 172.16.10.253

Is LAN1 having two stanby LAN cards while LAN 2 has none?

How can I fix this if so?

thanks!!

[Steven E. Protter]

Have both cards wired into the network/switch environment.

Plan a test whereby one of the primary non-heartbeat lan card is disconnected. Then try and access the server via the common (floating) IP address.

If you gain access, you know things work, if not, you have more work to do.

SEP

[Rajesh SB · Accepted solution]

Hi,

There is command to see the HA Cluster configuration details
# cmscancl -o /tmp/myscancl.log

To gather the configuration information from all the nodes in the
cluster and save the output to the file: /tmp/scancl.log

The output from this command contains the following information:

LAN device configuration (output from lanscan)

network status and interfaces (output from netstat)

file systems (output from mount)

LVM configuration (contents of /etc/lvmtab file)

LVM physical vg information (contents of /etc/lvmpvg file)

link-level connectivity (output from linkloop)

binary configuration file data
(output from cmviewconf)

Here you can clarify your all doubts.

Regards,
Rajesh

[Ivajlo Yanakiev]

you can use linkloop and check connection lan by lan using there MAC address

[melvyn burnard]

Normally, when a cluster is created, and the ascii file is written out after running cmquerycl, there SHOULD be comments below the network info per node, showing which lans could be a standby lan.
If this is missing, or you suspect it is incorrect, run cmquerycl again and specify a "test " file name, then check the comments in there.
The other way to tell what has actuaaly been confuigured in the cluster is to read the binary, and look to see which lans are on the same "bridged net".
Use cmscancl to get the best result, including linkloop tests, or just use the cmgetconf command, redirecting the output to a file