[Linux-HA] My Second Pacemaker Cluster

Robinson, Eric eric.robinson at psmnv.com
Wed Oct 20 05:57:19 MDT 2010


> You need a unique network socket for each ringnumber. At least that's 
> the recommendation if you want to use resource-level 
> fencing => http://www.drbd.org/users-guide/s-pacemaker-fencing.html

Thanks for your feedback, Dan. 

Now I'm a little worried because my existing cluster is clearly NOT
using a unique socket for each ringnumber, but it works fine. The
interface section looks like this...

        interface {
                ringnumber: 0
                bindnetaddr: 192.168.10.0
                mcastaddr: 226.94.1.1
                mcastport: 4000
        }

        interface {
                ringnumber: 1
                bindnetaddr: 198.51.100.0 
                mcastaddr: 226.94.1.1
                mcastport: 4000 
        }

I am using the same socket (mcastaddr & mcastport) for both ringnumber 0
and 1. Is this working because those two network segments are physically
separate and cannot see eachother?

I drew this really cool map before I saw your response and I hate to let
it go to waste, so let me use it to clarify. 


                        Exisiting 2-Node Cluster

                      |----(198.51.100.0/30)---|
                      |                        |
            |---------------------|  |---------------------|
            |        eth3         |  |        eth3         |
            |                     |  |                     |
            |     CLUSTER1_A      |  |     CLUSTER1_B      |
            |                     |  |                     |
            |   eth0       eth1   |  |   eth0       eth1   |
            |     |--bond0--|     |  |     |--bond0--|     |
            |          |          |  |          |          |
            |---------------------|  |---------------------|
                       |                        |
                       |                        |
----------------------------(192.168.10.0/24)---------------------------
----
	    |                      |                               |
          |                      |                               |
|-------------------|  |-------------------|
|-------------------|
|         |         |  |         |         |           |         |
|
|    |--bond0--|    |  |    |--bond0--|    |           |    |--bond0--|
|
|  eth0       eth1  |  |  eth0       eth1  |           |  eth0
eth1  |
|                   |  |                   |           |
|
|    CLUSTER2_A     |  |    CLUSTER1_2B    |           |    CLUSTER1_2C
|
|                   |  |                   |           |
|
|  eth2             |  |             eth3  |           |  eth3
eth2  |
|-------------------|  |-------------------|
|-------------------|
     |                                 |                     |         |
     |                                 |--(198.51.100.4/30)--|         |
     |                                                                 |
     |----------------(198.51.100.8/30)--------------------------------|

                                New 3-Node Cluster


So on CLUSTER2, I'm thinking the interface sections need to look like
this...


        interface {
                ringnumber: 0
                bindnetaddr: 192.168.10.0
                mcastaddr: 226.94.1.2
                mcastport: 4002
        }

        interface {
                ringnumber: 1
                bindnetaddr: 198.51.100.4 
                mcastaddr: 226.94.1.2
                mcastport: 4002
        }
        interface {
                ringnumber: 2
                bindnetaddr: 198.51.100.8 
                mcastaddr: 226.94.1.2
                mcastport: 4002 
        }


To me, this looks as if it should work fine. Does this look right to
anyone else?

[Observation: I can achieve a unique socket by varrying either the IP
address or the port. It sounds like you prefer to vary the IP and leave
the port unchanged. I prefer to use the same mcastaddr IP on all nodes
and vary the port. I don't know if there is a strong argument either
way.]

--
Eric Robinson


Disclaimer - October 20, 2010 
This email and any files transmitted with it are confidential and intended solely for General Linux-HA mailing list. If you are not the named addressee you should not disseminate, distribute, copy or alter this email. Any views or opinions presented in this email are solely those of the author and might not represent those of Physicians' Managed Care or Physician Select Management. Warning: Although Physicians' Managed Care or Physician Select Management has taken reasonable precautions to ensure no viruses are present in this email, the company cannot accept responsibility for any loss or damage arising from the use of this email or attachments. 
This disclaimer was added by Policy Patrol: http://www.policypatrol.com/



More information about the Linux-HA mailing list