Information Technology Reference
In-Depth Information
and use Storage vMotion. If your array supports dynamic growth of LUNs, you could
grow the VMFS or NFS datastores, and if it doesn't, you could add more VMFS extents.
Master It
Using the coni gurations in the previous question, what would the minimum
amount of raw capacity need to be if the VMs are actually only 20 GB of data in each VM,
even though they are provisioning 40 GB and you used thick on an array that didn't sup-
port thin provisioning? What if the array
did
support thin provisioning? What if you used
Storage vMotion to convert from thick to thin (both in the case where the array supports
thin provisioning and in the case where it doesn't)?
Solution
If you use thick virtual disks on an array that doesn't support thin provision-
ing, the answers are the same as for the previous question. If you use an array that does
support thin provisioning, the answers are cut down by 50 percent: 20 TB for RAID 10,
6.25 TB for RAID 5 (4+1), and 6 TB for RAID 6 (10+2). If you use Storage vMotion to con-
vert to thin on the array that doesn't support thin provisioning, the result is the same, just
as it is if you do thin on thin.
Master It
Estimate the number of spindles needed for 100 VMs that drive 200 IOPS each
and are 40 GB in size. Assume no RAID loss or cache gain. How many if you use 500 GB
SATA 7200 RPM? 300 GB 10K Fibre Channel/SAS? 300 GB 15K Fibre Channel/SAS?
160 GB consumer-grade SSD? 200 GB Enterprise Flash?
Solution
This exercise highlights the foolishness of looking just at capacity in the server
use case. 100 × 40 GB = 4 TB usable × 200 IOPS = 20,000 IOPS. With 500 GB 7200 RPM,
that's 250 drives, which have 125 TB raw (ergo non-optimal). With 300 GB 10K RPM,
that's 167 drives, which have 50 TB raw (ergo non-optimal). With 15K RPM, that's 111
drives with 16 TB raw (getting closer). With consumer-grade SSD, that's 20 spindles and
3.2 TB raw (too little). With EFD, that's 4 spindles and 800 GB raw (too little). The moral
of the story is that the 15K RPM 146 GB drive is the sweet spot for this workload. Note
that the extra space can't be used unless you can i nd a workload that doesn't need any
performance at all; the spindles are working as hard as they can. Also note that the 4 TB
requirement was usable, and we were calculating the raw storage capacity. Therefore, in
this case, RAID 5, RAID 6, and RAID 10 would all have extra usable capacity in the end.
It's unusual to have all VMs with a common workload, and 200 IOPS (as an average) is
relatively high. This exercise also shows why it's efi cient to have several tiers and several
datastores for different classes of VMs (put some on SATA, some on Fibre Channel, some
on EFD or SSD)—because you can be more efi cient.
Chapter 7: Ensuring High Availability and Business
Continuity
Understand Windows clustering and the different types of clusters.
Windows clustering
plays a central role in the design of any high-availability solution for both virtual and physi-
cal servers. Microsoft Windows clustering gives us the ability to have application failover to
the secondary server when the primary server fails.
Master It
Specii cally with regard to Windows clustering in a virtual environment,
what are three different types of cluster coni gurations that you can have?
