What's the problem exaclty??? ... Error: Detected a partition of IP subnet

[Enrico Venturi]

Checking cluster file: cmclconf.ascii
Checking nodes ... Done
Checking existing configuration ... Done
Gathering configuration information ... Done
Gathering configuration information ... Done
Gathering configuration information ............. Done

Warning: Volume group /dev/1353NM_1 is configured differently on node gemha1 than
on node gemha2
Warning: Volume group /dev/1353NM_1 is configured differently on node gemha2 than
on node gemha1
Error: Detected a partition of IP subnet 10.0.0.0.
Partition 1
gemha1 lan1
Partition 2
gemha2 lan1
Failed to evaluate network
cmcheckconf : Unable to reconcile configuration file cmclconf.ascii
with discovered configuration information.
gemha1,sys,root #

[G. Vrijhoeven]

Hi,

Did you subnet the 10.0.0.0 network other than 255.255.255.0?

If so i think this is your problem.

Gideon

[Karthik S S]

Hi,

Are you trying to configure the heartbeat LAN on the 10.0.0.0 network?. If yes on gemha2 have you configured any NIC in the 10.0.0.0 subnet?

Post cmclconf.asc. Also /etc/rc.config.d/netconf and the output of netstat -nr of both the nodes.

Regards,
Karthik S S

[Massimo Bianchi · Accepted solution]

Hi,
usually i got this error when the cabling of the network is not correct, because the cluster cannot resolve correctly the visibility.


The cluster tries to do a "linkloop" between every lan card with an ip address and the deputatted standby, and the cards on the other node on the same subnet, and they MUST be all reachable.

If you have some vlan, and the cabling is not correct, it can happen that the standby lan cannot linkloop the primary lan.

Check it.

HTH,
Massimo

[Mohanasundaram_1]

Hi Enrico,

have you done the import of VG using a map file?
1) Check if the VG 1353NM_1 is having the same number of disk devices in both the Nodes. Also create a map file and import to the other system.
# vgexport -p -s -m /tmp/1353NM_1.map /dev/1353NM_1

# rcp /tmp/1353NM_1.map

at the other node:
# vgimport -s -m /tmp/1353NM_1.map

2) Also verify if the Minor number for the group file is proper in both the servers.

As for the LAN error, I always do a test of linkloop from one server to the other before running cmcheckconf.
# linkloop -i 0
# linkloop -i 1

If this does not come as "OK", your cmcheckconf will not succeed.

Sometimes VLAN configuration may cause such problems even though the linkloop worked.

Check also the subnet mask for the LAN1 in both the servers.

All the best.

Rgds,
Mohan.