[stiven.munozMember]

Hi all,

I have a problem with the configuration of the cluster, i explain my configuration, all of the infrastructure is in VMware:

Node1Ip: 192.168.10.10
Node2Ip: 192.168.10.11
ClusterIp: 192.168.10.12

This configuration is ok, and works when i test the cluster configuration. But, when i add another eth, because i need to configure a farm in other network with other VLAN.
In this case i tried to modify in the configuration file “$exclude=”–exclude if_eth0_conf –exclude if_eth1_conf”;”, and configure

Node1eth1: 192.168.20.11
Node2eth1: 192.168.20.12
Virtual interface: 192.168.20.13

I create in the master the virtual interface, and it replicate to the backup. Here all is working, ip 20.13 responds to ping. But when i force balance to Node2, all seems works, but the ip doesn’t respond to ping, and obviously doesn’t work farm.

can you help me please? Is my first time with this software.

Thanks.

[JoseSupport]

Hi Stiven,

Welcome to Relianoid Community.

Could you please generate a support save (at System > Support Save) with each node and them at: support AT relianoid DOT com ?

Thank you.

[JoseSupport]

Hi Stiven,

I have reviewed the supportsave files but I need more information about the state of the node when the issue is happening.
Can you please share a new supportsave of the second node when the issue is happening?

If you can’t access the web interface at that time please run this command from the console: /usr/local/relianoid/bin/supportsave
Then, after restarting the node you can scp the new file located in /tmp.

Thank you.

[JoseSupport]

Hi Stiven,

Thanks for sending the requested information. I don’t see any issue on logs or configuration.

Please verify the MAC address of the floating virtual IPs are changing when the node 1 is primary and when the node 2 is primary. You can try for every IP address:

ping -c1 IP_ADDRESS
arp -n | grep IP_ADDRESS

Make sure changing the MAC address of IP addresses are allowed on the security settings of VMware virtual switch.

Let us know if you need more help or the issue was solved.

[stiven.munozMember]

I configured the Vmware virtual switch to allow mac changes for the 2 different vlan’s, in this case, the virtual interface asigned to cluster continuous working when the cluster node master changes by fail to the other, but the other virtual interfaces doesn’t work, and i don’t know if i have a mistake of configuration.

Thanks

[JoseSupport]

Hi Stiven,

I’ve seen your message:

I think i found the problem. When the failover occurs, the new master non announces to the gateway the new mac for ARP. When I execute this command “arping -U 172.16.212.1 -s 172.16.212.220 -c1” from the new master, the ARP of gateway is updated, and all works. ¿How can I do it automatically?

So, to automate this when a node becomes master is adding this command to the file /usr/local/relianoid/config/noid-start.
Make sure this change is made on both nodes.

I hope this helps.

[Author]

Posts