[openstack-hpc] VM fail over mechanism to Remote DR site

KK CHN kkchn.in at gmail.com
Fri Jul 9 17:03:37 UTC 2021


Members,
I am using OpenStack Ussuri for our datacentre for virtualization
requirements. we are using KVM hypervisor on Debian .

We are in a requirement that  our DR site is 200+ KMs apart from our office
premises and connectivity over WAN. This DR site also going to deploy with
Ussuri version.

How we can setup the DR site so effectively so that, when VMs  or Physical
hosts in our DC clusters   crashes or abruptly power to one physical
machine got break due to  electrical wire pulled out/unplugged  , then how
automatic failover to the DR site which is  300 miles away from our DC .

 what will be the solution for this setup so that automatically the VMs
crashed in our DC will be served from our DR site with minimum latency or
zero downtime (is zero down time possible ? kindly enlight on this or how
to achieve near zero down time ) using OpenStack without any data loss ( is
this possible no Data Loss ?). How ?

Whats the best architecture for this using openstack, what all are the
tools components need to be used ? and what type of solutions need to be
deployed for the above requirement.

VMs are nearly  100 nos and Linux and Windows majority of them with around
100 GB to 200 GB size for each VM.

Kindly share your thoughts and expertise in this regard and enlighten me on
what techniques / components of OpenStack need to be employed for achieving
this.

Any hints and tips  and reference links highly appreciated.

Thanks in advance,
Kris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-hpc/attachments/20210709/c8cef026/attachment.html>


More information about the OpenStack-HPC mailing list