[openstack-dev] Openstack disaster recover with CloudFerry, other tools?

Marzi, Fausto fausto.marzi at hpe.com
Thu Oct 29 21:01:07 UTC 2015


Hi Jonathan,
We are using Freezer for backup restore and disaster recovery (http://wiki.openstack.org/wiki/Freezer). It gives us flexibility as multiple storage media are supported (swift, ssh, local fs).
So for example you may want to use ssh, to recover in case keystone or swift are not available.

We are also working on parallel storage media for backups and restore. So the user can use 2 swift with independent credentials or ssh + swift and so on.

We are very actively involved on the development.  Please let us know if there's anything we can do for you here or on #openstack-freezer.

Many thanks,
Fausto

-----Original Message-----
From: Jonathan Brownell [mailto:cadenzajon at gmail.com] 
Sent: 23 October 2015 20:04
To: openstack-dev at lists.openstack.org
Subject: [openstack-dev] Openstack disaster recover with CloudFerry, other tools?

Hello, I'm interested in using technology like CloudFerry [https://github.com/MirantisWorkloadMobility/CloudFerry] to migrate OpenStack resources from one cloud to another in the use case of disaster recovery.

I can deal with the storage replication necessary to make sure that the storage backend(s) are regularly freshened in the failover cloud, and its files will just need to be reattached to Cinder volume and Glance image objects during reconstruction (in preparation for association with new, failed-over compute instances).

CloudFerry is designed to migrate resources from one cloud to another while both environments are accessible and operable (i.e. its primary "Openstack version upgrade" scenario). So, for my use case, I expect to have to define metadata that would be regularly collected (via APIs and DB), transmitted, and cached on the failover side in order to perform a recovery if the primary cloud goes completely offline.

I can see a number of OpenStack Summit presentations over the years that describe this kind of method for failing over resources from one cloud to another to address disaster recovery, but have not found any other projects or tools that help accomplish this. Is there work in the community that targets this kind of functionality today that I should familiarize myself with? Or any huge red flags I'm missing that would prohibit this kind of solution from working?

Thanks,

-JB

__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev



More information about the OpenStack-dev mailing list