That covers redundancy in case of a network card or cable failure, but isn't what partitioning is about. Multiple NICs cannot prevent partitioning.

As an example, partitioning might happen when a network switch between two network segments fails so you have nodes A+B in segment 1 able to talk to each other but unable to talk to nodes C+D in segment 2, while C+D can talk to each other but not A+B.

Pacemaker/corosync contain a lot of algorithms to fence off partitions without quorum and can resort to things like STONITH if required to force a node to shutdown rather than risk it causing disruption to the cluster (for example if it tries to take over traffic to a virtual IP you could end up in a case where you have two servers sending ARP responses for the same IP).

-Steve



On 10 February 2013 15:27, Eliot Gable <egable+freeswitch@gmail.com> wrote:
You use multiple NICs in the systems and send heartbeats out all of
them. There is no other way to do it. Two NICs are required, and 3 are
recommended.


On Sun, Feb 10, 2013 at 6:54 AM, Marcin Gozdalik <gozdal@gmail.com> wrote:
> I wonder how do you handle network partition? Simply "knowing" which nodes
> are up in a distributed system is hard enough and there is a reason why
> solutions like pacemaker/corosync have years of development behind them.
>
> W dniu niedziela, 10 lutego 2013 użytkownik Eliot Gable napisał:
>>
>> The beauty of the module is that it will not require any 3rd party
>> software to operate. I am designing it to run entirely in-memory using
>> multi-cast events to communicate between cluster nodes. No external
>> database will be required to enable the automatic fail over and call
>> recovery.
>>
>> If you have your customers / clients using DNS to point traffic to
>> your system and you run multiple master nodes with DNS performing load
>> balancing, you can also make use of the maintenance mode I have
>> planned which will use a separate IP address and update your DNS
>> records to remove the master which you put into maintenance mode and
>> put in the maintenance IP which is assigned to the slave node which is
>> being promoted as master. This allows calls to drain off the master
>> while new calls come up on the slave, and then you can gracefully shut
>> down the master when all calls are finished. Once you are done with
>> maintenance, you simply bring it back online and it comes up as a
>> slave or a master as needed (according to your configured master /
>> slave ratio). The IP which it was using as a master node becomes the
>> new designated maintenance IP. This is the only case where a 3rd party
>> software package is required, and it is an optional feature.
>>
>>
>> On Sat, Feb 9, 2013 at 6:49 PM, Gilad Abada <ga@steadfasttelecom.com>
>> wrote:
>> > Hey Eliot
>> >
>> > Sounds exciting. Which DB will you be using Postgres? I will try to
>> > get my company to pitch in a few dollars.
>> >
>> > Sent from my mobile device.
>> >
>> > On Feb 9, 2013, at 12:11 PM, Eliot Gable <egable+freeswitch@gmail.com>
>> > wrote:
>> >
>> >> For the past two years, I have been working on and off on a high
>> >> availability cluster module for FreeSWITCH. If you would like to see
>> >> HA built into FreeSWITCH directly, instead of dealing with Pacemaker +
>> >> Corosync or other 3rd party systems, please check out the module,
>> >> voice your support, provide feedback and suggestions, and consider
>> >> supporting it.
>> >>
>> >> Some advantages over a 3rd party solution like Pacemaker include:
>> >>
>> >> - Sub-second failure detection and response (meaning it can start
>> >> recovering calls less than a second after a failure occurs)
>> >> - Failure detection is built into FreeSWITCH so it can detect
>> >> conditions which 3rd party modules cannot detect
>> >> - One shared configuration file between all nodes (no per-node IP
>> >> configuration required)
>> >> - No fixed number of nodes
>> >> - Graceful migration of calls with a maintenance mode (requires DNS)
>> >>
>> >> These are just some of the benefits of building a FreeSWITCH-specific
>> >> HA system. For additional details about the project, please visit:
>> >>
>> >> http://wiki.freeswitch.org/wiki/Mod_ha_cluster
>> >>
>> >>
>> >> If you would like to see this project come to fruition, please
>> >> consider financially supporting it through Kickstarter:
>> >>
>> >>
>> >> http://www.kickstarter.com/projects/1175121367/high-availability-cluster-module-for-freeswitch
>> >>
>> >>
>> >>
>> >> Thanks in advance for anyone who contributes, either with just a
>> >> shout-out, with feedback and/or suggestions, and especially to those
>> >> who pledge some financial support.
>> >>
>>
>> _________________________________________________________________________
>> Professional FreeSWITCH Consulting Services:
>> consulting@freeswitch.org
>> http://www.freeswitchsolutions.com
>>
>> FreeSWITCH-powered IP PBX: The CudaTel Communication Server
>> http://www.cudatel.com
>>
>> Official FreeSWITCH Sites
>> http://www.freeswitch.org
>> http://wiki.freeswitch.org
>> http://www.cluecon.com
>>
>> FreeSWITCH-users mailing list
>> FreeSWITCH-users@lists.freeswitch.org
>> http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
>> UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
>> http://www.freeswitch.org
>
>
>
> --
> Marcin Gozdalik
>
> _________________________________________________________________________
> Professional FreeSWITCH Consulting Services:
> consulting@freeswitch.org
> http://www.freeswitchsolutions.com
>
> FreeSWITCH-powered IP PBX: The CudaTel Communication Server
> http://www.cudatel.com
>
> Official FreeSWITCH Sites
> http://www.freeswitch.org
> http://wiki.freeswitch.org
> http://www.cluecon.com
>
> FreeSWITCH-users mailing list
> FreeSWITCH-users@lists.freeswitch.org
> http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
> UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
> http://www.freeswitch.org
>



--
Eliot Gable

"We do not inherit the Earth from our ancestors: we borrow it from our
children." ~David Brower

"I decided the words were too conservative for me. We're not borrowing
from our children, we're stealing from them--and it's not even
considered to be a crime." ~David Brower

"Esse oportet ut vivas, non vivere ut edas." (Thou shouldst eat to
live; not live to eat.) ~Marcus Tullius Cicero

_________________________________________________________________________
Professional FreeSWITCH Consulting Services:
consulting@freeswitch.org
http://www.freeswitchsolutions.com

FreeSWITCH-powered IP PBX: The CudaTel Communication Server
http://www.cudatel.com

Official FreeSWITCH Sites
http://www.freeswitch.org
http://wiki.freeswitch.org
http://www.cluecon.com

FreeSWITCH-users mailing list
FreeSWITCH-users@lists.freeswitch.org
http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
http://www.freeswitch.org