As of DDOS 6.0 there is a new feature known as long term retention (LTR). This allows you to add a tier of cloud (object) storage to a DDR then migrate a subset of data from locally attached disks (the active tier) to the cloud. LTR supports various cloud providers such as AWS/Azure/Virtustream/ECS so you can use private or public clouds.
A couple of things to note:
- All backups must first hit the active tier and age at least 14 days before they can be migrated to the cloud
- You cannot read backups directly from the cloud - corresponding files have to be 'recalled' to the active tier before they can be read
If you take a look at the DDOS 6.0 administration guide this will contain more details around LTR.
The only issue, however, is that by default you would still have only one copy of your data and be reliant on a single DDR (whether the data was on local or cloud storage). If you were creative with fastcopies e.t.c. you could create two copies of data - one on locally attached storage and another on the cloud however the issue with a single DDR (hence single point of failure) remains. To be honest if you need completely separate copies theres not much you can do apart from have two DDRs and replicate data between them.
James_Ford
30 Posts
0
March 31st, 2017 06:00
Hi jshweg,
As of DDOS 6.0 there is a new feature known as long term retention (LTR). This allows you to add a tier of cloud (object) storage to a DDR then migrate a subset of data from locally attached disks (the active tier) to the cloud. LTR supports various cloud providers such as AWS/Azure/Virtustream/ECS so you can use private or public clouds.
A couple of things to note:
- All backups must first hit the active tier and age at least 14 days before they can be migrated to the cloud
- You cannot read backups directly from the cloud - corresponding files have to be 'recalled' to the active tier before they can be read
If you take a look at the DDOS 6.0 administration guide this will contain more details around LTR.
The only issue, however, is that by default you would still have only one copy of your data and be reliant on a single DDR (whether the data was on local or cloud storage). If you were creative with fastcopies e.t.c. you could create two copies of data - one on locally attached storage and another on the cloud however the issue with a single DDR (hence single point of failure) remains. To be honest if you need completely separate copies theres not much you can do apart from have two DDRs and replicate data between them.
Thanks, James