Introduction
You probably have read by now that is Storage Replica will be available on Windows Server 2019 Standard Edition. Until now with Windows Server 2016 and 2019 Previews it was a Datacenter Edition only feature. If you haven’t go read up on it over at
Announcing Windows Server 2019 Insider Preview Build 17639. This is good news and I’m very curious to see what use cases and solutions people will leverage this capability for.
Storage Replica on Standard Edition has 2 major limitations:
- SR replicates a single volume (instead of an unlimited number of volumes).
- Volumes can have one partnership (instead of an unlimited number of partners). Volumes can have a size of up to 2 TB (instead of an unlimited size).
My first use case
This is why I’m a bit disappointed right now as I had hoped it would help me with some data protection use cases. As part of my roles and expertise lies in building creative, cost effective data protection solutions adhering to at least the 3-2-1 rule in both on premises as well as in private, hybrid and public cloud setting or any combination of the above.
On of the challenges is getting backup data off site. One solution is to replicate the backups. This mostly doesn’t require synchronous replication as most of the time the bandwidth isn’t there or there is no need. Many cost effective storage solutions don’t have storage replication or it comes a significant license cost. The backup software often has this but it works a per definition less effective layer than Storage Replication.
Cost effective but effective backup targets don’t always need to be high available, that depends on the needs and you can cluster a backup target file server role with Windows Server standard edition.
So for my use case I would rather have seen asynchronous only in Windows Server 2019 Standard Edition but without the volume size and number limitation and be done with it.
This would give me off site data protection of my backups via storage Replica which is tremendously efficient. It doesn’t rely on backup software or other solution operating at the file level and as such that are inherently less efficient. But backups often mean more then one target volume and larger volume sizes than 2TB. The beauty of Storage Replica is that it’s completely storage agnostic solution and we can build the solution on top of whatever is at hand.
Some data protection use cases where the limits might not matter
With the current limits it might fit in with some SOHO/ ROBO scenario’s perhaps. When you have 10 branch offices with Standard Edition you could potentially replicate a volume to a central datacenter edition for safe keeping off site. But in those scenarios we’re also looking at Azure File Sync and offload the data protection to the cloud if/when possible.
Also remember that “pure” MPLS isn’t the only answer anymore to many connectivity challenges anymore but SDWAN/local Internet breakouts are eating part of their cake . MPLS costs versus VPN is also a reality which limits bandwidth (let’s face it, it’s a cost issue), perhaps further reducing desirability Storage Replica for this use case. SD-WAN could be helping address “old school” network cost optimization limitations for this use case by delivering more and better than VPN without the need for MPLS or express route etc. Just thinking out loud here. It’s a cloud first world, where servers still have a role to play, but for that they need to be flexible and allow for many possible permutations. It’s a missed opportunity I’d say. My opinion, but there is hope. Microsoft states the following on Storage Replica Standard
So if my use case make enough sense to them they might change something still but we’ll see. Anyway, my 5$ cents on the subject.