cmapplyconf don''t work for lan fail

[claudio_22]

Hello ,

I want configure a lan like standby in a cluster, with node1 and node2 ; but cmcheckconf fail because this lan on node2 has not a valid ip .. and so on ..

This outuput of lanscan on node2

Path Address In# State NamePPA ID Type Support Mjr#
1/0/1/1/0/4/0 0x000F202B1508 0 UP lan0 snap0 1 ETHER Yes 119
1/0/6/1/0 0x001321BDFD6E 1 UP lan1 snap1 2 ETHER Yes 119
1/0/8/1/0 0x001321BDFD4D 2 UP lan2 snap2 3 ETHER Yes 119
1/0/10/1/0 0x001321BDFD3F 3 UP lan3 snap3 4 ETHER Yes 119
1/0/12/1/0 0x001321BDFD4B 4 UP lan4 snap4 5 ETHER Yes 119
1/0/14/1/0 0x001321BDFD52 5 UP lan5 snap5 6 ETHER Yes 119


linkloop on lan2 that i want use ..

linkloop 0x001321BDFD4D
Link connectivity to LAN station: 0x001321BDFD4D
error: get_msg2 getmsg failed, errno = 4
-- FAILED
frames sent : 1
frames received correctly : 0
reads that timed out : 1

Linkloop also fail on lan3/4 and 5 .

On node1 i've same lan situation .. but lan2 linkloop is ok while others fails too.


The cluster is on 2 ia64 hp server rx7620
with hp-ux 11.23 patch bundle may2005

I'm thinking at an harware failure . What is strange is that if I go in lanadmin on node2 and select lan2 is how me is up .. ?!

Thanks for any suggest . These system should go in production in few days

[melvyn burnard]

You appear to have a network cabling/configuration issue here.
This must be corrected for cmquerycl to correctly pick and use the various lan cards, both as primary (configured with IP address) an dstandby ( completely unconfigured).
Until you sort out the connectivity issues, cmquerycl/cmcheckconf/cmapplyconf will not work correctly

[claudio_22]

Hi ,

I've run following command

lipofl01:root/home/root# cmquerycl -k -w full -c PCP


Node Names: lipodb01
lipofl01

Bridged networks (full probing performed):

1 lan0 (lipodb01)
lan2 (lipodb01)
lan0 (lipofl01)

2 lan1 (lipodb01)
lan1 (lipofl01)

IP subnets:

IPv4:

172.17.15.0 lan0 (lipodb01)
lan0 (lipofl01)

192.168.16.16 lan1 (lipodb01)
lan1 (lipofl01)

IPv6:

Possible Heartbeat IPs:

172.17.15.0 172.17.15.32 (lipodb01)
172.17.15.33 (lipofl01)

192.168.16.16 192.168.16.17 (lipodb01)
192.168.16.18 (lipofl01)


lan2 is not seen by lipofl01

[melvyn burnard]

This shows you have a connection issue on your network prventing the lan from being seen by both nodes. Investigate the cableing and switch configurations

[claudio_22]

Thanks ..but then why in lanadmin I see operation is up ? Again the lan is attached to a switch ; we have used another cable and another port of the switch but with no success .

So , don't seem to be lan card .. don't seem to be cable or switch port ..

Then , what could be ?

[Mark Nieuwboer]

Hi,

When linkloop fails and you have a cable in the lancard.
The switch isn't configured correct.
Look at your switch which ip-adress it;s get and which ip-adress you have given it on your system.

grtz. Mark

[melvyn burnard · Accepted solution]

the lanadmin just shows that the card is up, and appears to have a link out to the switch.
The fact you cannot linkloop to it show syou have a connectivity or configuration issue

[claudio_22]

So for you the link to the switch is ok ..

What type of configuration check can I do ?

[claudio_22]

Problem was on switch . That MAC address wasn't on the same VLAN that others

Thanks