We've recently updated our master\media appliance from 3.1 to 4.x with a view to introducing a cloud storage server as this version no longer needs a dedicated cloud catalyst server. As part of this configuration I believe I may need to specify a location for the CC cache?
Is there a default location for this? I thought I read that it actually uses the MSDP as the cache location? or perhaps creates a sub-directory called 'upload' but just a little uncertain to what I should specify here.
Has anyone done this on an appliance already and what settings were used?
thanks in advance
Solved! Go to Solution.
You are correct the media server will utilise some of the existing MSDP storage for the cache. The size requirement for this cache have reduced substantially so the usage will not be noticed. You do not need to specify this cache location assuming you already have an existing MSDP volume on the server (in fact even if you only create a cloud LSU, I don't think you need to specify the location - it will do the necessary directory creation).
Refer to the Deduplication Guide for some of the details on the parameters in the cloud.json, contentrouter.cfg and spa.cfg files on the storage server for details on the cache settings.
That's helped enormously - and I've taken a closer look at the 9.1 dedupe guide to clarify a couple of other points. There is still a couple of fundamental questions I have with Azure Archive tier with 9.1. Perhaps you can answer these?
Firstly - Does 9.1 still need to transition the backup images through an Azure Hot tier before moving into Azure Archive? I thought I saw some marketing information positioning 9.1 as 'Direct to Azure Archive'
Secondly - Does Azure archive support deduplicated data or do we need to calculate the costs of storing backup images in the Azure archive tier based on the hydrated size of the image? The NetBackup 9.1 cloud admin guide says "Accelerator and deduplication isn't supported for Azure Archive" whereas the 9.1 dedupe guide says it can be used as a optimised dedupe target but not a source. This is a little confusing.
Should I be reading this as Azure Archive can host deduplicated images but I can't target a dedupe backup directly to the Azure Archive LSU, I first need to backup to a local MSDP and then use a lifecycle to duplicate the data (optimised duplication) to the Azure Archive tier?
Firstly my reading is that the backup data and metadata are sent initially to the hot tier, then the data only is migrated to the archive tier. Accelerator is not supported for backups sent to this LSU type. The marketing that it is "direct to archive" I think refers to being able to send a backup directly to this LSU (the fact it transitions via the hot tier on the way is merely the mechanism it gets there).
Secondly, it most certainly does support deduplicated data. What is not supported is using this data as the source for an optimised duplication or in fact any operation other than restore (and note the considerations mentioned on this).
Hope this helps
I've another appliance that's running 18.104.22.168
I understand that MSDP Cloud was introduced in 8.3 so should be available to me on this appliance also. Is everything we've discussed around dedupe on the archive tier supported in this release or must I go to 4.1 to get these features?