Website under construction

Scenario 4: Storage node restart or maintenance 
In this scenario, Storage Spaces does not automatically retire physical drives from the storage pool for 
the reasons described earlier in scenario 3. When the storage node comes back online, Storage 
Spaces automatically updates all extents that are not up to date with the copies that were unaffected 
by the restart or maintenance. 
Scenario 5: Permanent storage node failure 
In this scenario, Storage Spaces requires the administrator to retire all of the affected physical drives 
from the storage pool, add additional storage nodes to the storage system if needed, and then begin 
repair. This is not an automatic process, because Storage Spaces does not know if the failure is 
temporary or permanent. It is not desired to initiate a repair that could potentially result in significant 
repair activity. 
More info To learn more about Storage Spaces Direct in Windows Server 2016, go to 
https://technet.microsoft.com/library/mt126109.aspx
. 
Deduplication 
In Windows Server 2016 Deduplication, the core focus is to have significant impact on the scale and 
performance that you can address with this technology. With this shift in focus, you can now use 
Windows Server 2016 in the following scenarios: 

Volumes up to 64 TB 

File sizes up to 1 TB 

Virtualized backup 

---

60 
CHAPTER 2 | Software-defined datacenter 

Nano server support 

Rolling cluster upgrades 
Since deduplication was introduced in Windows Server 2012, the core principles for getting the data 
to a chunked state have remained the same, now let us address what has changed that makes these 
new scenarios possible. 
The optimization engine has been upgraded from single thread to multiparallel thread using multiple 
I/O streams, as shown in Figure 2-62.