You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@ignite.apache.org by Vj Anand <vj...@automationanywhere.com> on 2018/11/17 15:39:07 UTC

Network partitioning

Hi Ignite Developers,

Here is the scenario that I am trying to  mitigate – I have a two node ignite cluster as cache; I am using Partitioning as the cache policy. I need to address network partitioning and avoid the split-brain scenario. What are the options?
Are there hooks into the Ignite cache that I can use to be notified when, there is a partition, or when nodes rejoin? Any pointers? Code Snippets? I am trying to avoid external quorum mechanism.

Thanks
VJ

Re: Network partitioning

Posted by Denis Magda <dm...@apache.org>.

Hi,

Ignite comes with partition loss policies that might be useful:
https://apacheignite.readme.io/docs/partition-loss-policies

As for the network segmentation, check out GridGain that provides this for
Ignite users.

Denis


On Sat, Nov 17, 2018 at 7:39 AM Vj Anand <vj...@automationanywhere.com>
wrote:

> Hi Ignite Developers,
>
> Here is the scenario that I am trying to  mitigate – I have a two node
> ignite cluster as cache; I am using Partitioning as the cache policy. I
> need to address network partitioning and avoid the split-brain scenario.
> What are the options?
> Are there hooks into the Ignite cache that I can use to be notified when,
> there is a partition, or when nodes rejoin? Any pointers? Code Snippets? I
> am trying to avoid external quorum mechanism.
>
> Thanks
> VJ
>