[PVE-User] 3 node PVE cluster, need recommendation.

[admin at extremeshok.com]

Look at glusterfs and sheepdog

Sent from my iPhone 5

> On 05 Oct 2013, at 12:28 AM, Fábio Rabelo <fabio at fabiorabelo.wiki.br> wrote:
> 
> Some time ago I give it a try, and it is not reliable, after several
> Split-brain situations ... I gived-up ...
> 
> 
> Fábio Rabelo
> 
> 2013/10/4 Muhammad Yousuf Khan <sirtcp at gmail.com>:
>> does anyone have use this DRBD 3 Node cluster with Proxmox cluster.
>> 
>> http://www.drbd.org/users-guide/s-three-nodes.html
>> 
>> 
>> On Fri, Oct 4, 2013 at 11:42 PM, Muhammad Yousuf Khan <sirtcp at gmail.com>
>> wrote:
>>> 
>>> we are planning to host a SaaS for our few New customers.
>>> DRBD worked well for us but for 2 nodes. but what if Site goes down.we  we
>>> need things up and running with out single transaction lost into our
>>> database.
>>> 
>>> we also have drbd option but unfortunately DRBD till 8.x only support 2
>>> nodes.
>>> secondly drbd protocol A which is highly adviced for WAN, has some chances
>>> loosing data. as it work on TCP threshold.
>>> 
>>> as per my personal experience i thought first that we should use 2 node
>>> cluster PVE. then ZFS snapshot should be transferd to Remote site. since
>>> snapshots are not realtime replication, therefore data will be loss which we
>>> do not want. any advice where i can run 2 server on site and 1 server off
>>> site with no single transaction lost.
>>> 
>>> 
>>> Note : we will be hosting a Microsoft platform.
>>> 
>>> 
>>> any help would be highly appreciated.
>>> 
>>> Thanks,
>> 
>> 
>> 
>> _______________________________________________
>> pve-user mailing list
>> pve-user at pve.proxmox.com
>> http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user
> _______________________________________________
> pve-user mailing list
> pve-user at pve.proxmox.com
> http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user