###BeginCaseStudy###
Case Study: 4
A Datum Corporation
Overview
A Datum Corporation is an accounting company.
The company has a main office and two branch offices. The main office is located in Miami.
The branch offices are located in New York and Seattle.
Existing Environment
Network Infrastructure
The network contains an Active Directory domain named adatum.com. All servers run
Windows Server 2008 R2. The main office has the following servers and client computers:
• Two domain controllers configured as DNS servers and DHCP servers
• One file server that has multiples shares
• One thousand client computers that run Windows 7
Each branch office has the following servers and client computers:
• One domain controller configured as a DNS server and a DHCP server
• Five hundred to 800 client computers that run Windows XP
Each office has multiple subnets. The network speed of the local area network (LAN) is 1
gigabit per second. The offices connect to each other by using a WAN link. The main office
is connected to the Internet.
Current Issues
The WAN link between the Miami office and the Seattle office is a low bandwidth link with
high latency. The link will not be replaced for another year.
Requirements
Application Requirements
The company is developing an application named App1. App1 is a multi-tier application that
will be sold as a service to customers.
Each instance of App1 is comprised of the following three tiers:
• A web front end
• A middle tier that uses Windows Communication Foundation (WCF)
• A Microsoft SQL Server 2008 R2 database on the back end
Each tier will be hosted on one or more virtual machines. Multiple-tiers cannot coexist on the
same virtual machine.
When customers purchase App1, they can select from one of the following service levels:
• Standard: Uses a single instance of each virtual machine required by App1. If a virtual
machine becomes unresponsive, the virtual machine must be restarted.
• Enterprise: Uses multiple instances of each virtual machine required by App1 to
provide high-availability and fault tolerance.
All virtual hard disk (VHD) files for App1 will be stored in a file share. The VHDs must be
available if a server fails.
You plan to deploy an application named App2. App2 is comprised of the following two
tiers:
• A web front end
• A dedicated SQL Server 2008 R2 database on the back end
App2 will be hosted on a set of virtual machines in a Hyper-V cluster in the Miami office.
The virtual machines will use dynamic IP addresses. A copy of the App2 virtual machines
will be maintained in the Seattle office.
App2 will be used by users from a partner company named Trey Research. Trey Research has
a single Active Directory domain named treyresearch.com. Treyresearch.com contains a
server that has the Active Directory Federation Services server role and all of the Active
Directory Federation Services (AD FS) role services installed.
Planned Changes
A Datum plans to implement the following changes:
• Replace all of the servers with new servers that run Windows Server 2012.
• Implement a private cloud by using Microsoft System Center 2012 to host instances
of App1.
• In the Miami office, deploy four new Hyper-V hosts to the perimeter network.
• In the Miami office, deploy two new Hyper-V hosts to the local network.
• In the Seattle office, deploy two new Hyper-V hosts.
• In the Miami office, implement a System Center 2012 Configuration Manager
primary site that has all of the system roles installed.
• Implement a public key infrastructure (PKI).
Notification Requirements
A Datum identifies the following notification requirements:
• Help desk tickets must be created and assigned automatically when an instance of
App1 becomes unresponsive.
• Customers who select the Enterprise service level must receive an email notification
each time a help desk ticket for their instance of App1 is opened or closed.
Technical Requirements
A Datum identifies the following technical requirements:
• Minimize costs whenever possible.
• Minimize the amount of WAN traffic
• Minimize the amount of administrative effort whenever possible.
• Provide the fastest possible failover for the virtual machines hosting App2.
• Ensure that administrators can view a consolidated report about the software updates
in all of the offices.
• Ensure that administrators in the Miami office can approve updates for the client
computers in all of the offices.
Security Requirements
A Datum identifies the following security requirements:
• An offline root certification authority (CA) must be configured.
• Client computers must be issued certificates by a server in their local office.
• Changes to the CA configuration settings and the CA security settings must be
logged.
• Client computers must be able to renew certificates automatically over the Internet.
• The number of permissions and privileges assigned to users must be minimized
whenever possible.
• Users from a group named Group1 must be able to create new instances of App1 in
the private cloud.
• Client computers must be issued new certificates when the computers are connected
to the local network only.
• The virtual machines used to host App2 must use BitLocker Drive Encryption
(BitLocker).
• Users from Trey Research must be able to access App2 by using their credentials
from treyresearch.com.
###EndCaseStudy###
You are evaluating the deployment of a multi-site Hyper-V failover cluster in the Miami office
and the Seattle office to host App2.
You need to identify which changes must be made to support the use of the multi-site
cluster.
Which changes should you identify?
A.
Configure all of the virtual machines to use dynamic memory. Implement Distributed File
System (DFS) Replication and store the virtual machine files in a replicated folder.
B.
Implement Distributed File System (DFS) Replication and store the virtual machine files in
a replicated folder. Upgrade the WAN link between the Miami and Seattle offices.
C.
Purchase a storage solution that replicates the virtual machines. Configure all of the
virtual machines to use dynamic memory.
D.
Upgrade the WAN link between the Miami and Seattle offices. Purchase a storage
solution that replicates the virtual machines.
https://technet.microsoft.com/en-us/library/dd197575(v=ws.10).aspx
D
0
0
MS rarely suggests a third party solution and not necessary in this scenario. I’d go with “B” except the Miami to Seattle WAN will not be replaced for another year. So “A”, but why? All BAD answers, choose the least incorrect one. “B”.
0
0
You must have a Fast WAN for Multisite clustering and DFS doesn’t work for multisite Hyper-V
Clusters
0
0
Agreed
https://technet.microsoft.com/en-us/library/dd197575(v=ws.10)
http://blogs.technet.com/b/meamcs/archive/2013/11/09/microsoft-windows-multi-site-failover-cluster-best-practices.aspx
So the answer must be C because the case states the WAN link won’t be replaced for 1yr.
0
0
Agreed with Paris & Workout.
I’ve done this (in a slightly different way) in the real world. This answer fits with reality and best practices.
0
0
I agree with paris, Workout and jimilives.
But the question is quite clear:
“You need to identify which changes must be made to support the use of the multi-site
cluster”, regardless of the actual WAN link between the Miami office and the Seattle office, which has low bandwidth link with high latency and will not be replaced for another year.
Therefore the ONLY right answer to the question is “D”
0
0
This question confuses me. I get that you need a fast WAN for multi site clustering but the case study quite clearly says that the “WAN link won’t be replaced for 1yr” so how can the answer be D?
0
0
D
From the technical requirements:
Provide the fastest possible failover for the virtual machines hosting App2.
This will necessitate upgrading the WAN link.
0
0
what about C ?
0
0
I’d go with D, because C makes no sense whatsoever… Whats the point of configuring VM’s to use dynamic memory? How will that ensure faster failover? It wont. But I don’t understand why the say that they won’t upgrade links for 1 more year… Also, answer C violates “minimize cost” I guess Storage solution costs more then a WAN link… But this question is retarded 101%
0
0
D seems best choice
Nobody seems to have taken into account synchronous v asynchronous replication
Asynchronous will work in the current infrastructure with low bandwidth/high latency but has data loss implications in the case of a disaster
With this in mind;
You are asked to identify which changes must be made to support the use of the multi-site cluster.
As A B & C are not acceptable (No DFS replication in multisite clusters)or just don’t make sense
The only option is D. You identify that the WAN link must be upgraded to support the multisite cluster without risk of data loss and a third party storage solution that includes replication technology
Gregg Shields is a King 🙂
http://www.backupacademy.com/multi-site-clustering-hyper-v-disaster-recovery.html
check around the 12:45 mark for synchronous/asynchronous replication but watch the whole thing to gain real knowledge
0
0