- Integrated Systems
- About Us
- Integrated Systems
- About Us
02-18-2018 07:31 AM
SimpliVity 380 Backup
With SimpliVity 380 Gen 10 and OmniStack version 3.7.1 I have several questions about native SimpliVity backup.
1-How exactly the SimpliVity backup restore works:
1.1-It is granular like Veeam? If yes, it needs some agent, script or some several steps do achieve?
1.2-The restored data goes back to where? To the same host that lost the data? The VM can be up and running during the restore job?
1.3-Is it possible to do the granular restore by delegating this task to a service desk without it having access to the contents of the backup?
1.4-There is a backup capacity planning tool?
04-06-2018 09:03 AM - last edited on 04-08-2018 11:58 PM by Parvez_AL
Re: SimpliVity 380 Backup
Sorry for the delay in answering this @mogicrz.
1 - In HPE SimpliVity, a backup is a first-class object and in every way a copy of the parent VM. If you have a backup you need to restore, you can restore it locally and overwrite the existing VM, restore in parrallel using a new name, or restore remotely. The only thing to bear in mind is that a remote restore may take some time if the data is not already present, as this data will need to be sent to the destination cluster.
2 - By granular, do you mean incremental? If so, no, it is not a delta in the traditional sense and as above, is a full copy of the parent VM in every way. The key difference is that it is deduplicated, which is functionally the same as a delta (for ease of comparison). If the blocks exist on the destination, a counter is simply incremented and no data is transmitted. Only if the data is unique is the data transmitted.
3 - As per 1. you can choose how to restore the VM. Regarding where, it depends if it is local or remote. Backups share the same data locations as the parent to maximise deduplication, but every object has n+1 failures to tolerate, so in a single failure, there is always a copy of the data available. Indeed, the VM itself is protected in the same way. NOTE: This is per object, not per cluster, so it is functionally different to other HCI offerings in this regard. If absolute protection is key for you, this is where remote backups come in. This will ensure that a copy of your VM is in a discrete failure domain, which may be another rack', another room in the same site, or another continent.
4 - Unfortunatly no. Restoring the VM (or the files within the VM) requires permissions over the object, so delegation is not possible here. It must be done by an environment administrator.
5 - As the system is primary storage, the 'workload' consists of the VM and it's backups. This then begs the question, would you want to seperate this out in to discrete components? In this paradigm shift, it probably makes more sense to think of the workload holistically, and manage the storage in terms of the common pool of storage. In this regard, everything is exposed in the vCenter GUI, but you can also conume this in a 3rd-party tool if you wish, via the API.
I hope this answers your questions!