MCSA 70-740 Flashcards
Potential indicators of a processor bottleneck
A sustained queue of greater than 1.
A sustained time of 85% or greater.
Normal disk queue length and % time
Queue of 2 plus the number of disk spindles.
Under 90%.
Potential indicator of a memory bottleneck
Over 20 pages per second
VHD is supported by what Generation(s) of VMs?
Virtual Disk supported by both Gen 1 and 2
VHDX is supported by what Generation(s) of VMs?
Virtual Disk supported by Gen 2 VMs only
Disk(s) and file system(s) compatible with Data Deduplication
Compatible with either MBR or GPT disks.
Compatible with NTFS only.
For a Windows Server Backup, what is contained in the “System State”?
- boot files
- Active Directory files
- sysvol, when run on a domain controller
- the registry
- other info, depending on the computer configuration and installed roles
Recommended block size for a disk containing SQL Server database and log files
64KB
Maximum size of a VHD?
2 TB
The default number of max failures before a role will transfer to a new node
1
What role do we add in Failover Cluster Manager to replicate virtual machines?
Replica Broker
Requirements for expanding or shrinking a virtual hard disk
- VHDX format (VHDs cannot be resized)
- SCSI controller (IDE is not supported)
- Dynamic type (Fixed disks cannot be resized)
What must be done in Active Directory for a Kerberos constrained delegation of a shared-nothing live migration?
The following services must be added to the Delegation tab for each server participating:
- cifs (used for the storage migration)
- Microsoft Virtual System Migration Service (used for the live migration)
True or False: In order for Microsoft to support a Failover Cluster, you must run all tests in the “Validate a Configuration Wizard” and have all run as successful.
True
When configuring a new failover cluster with a disk witness, if using the GUI wizard, what disk is used as the witness?
It will automatically choose the LUN shared by all nodes which is of the smallest size.
What are the types of quorum for a failover cluster?
- Disk Witness
- File Share Witness
- Cloud Witness
Define Guest Clustering and list some example usage scenarios
A Guest Cluster is a failover cluster consisting of VM nodes which themselves exist on a host cluster. Good usage scenarios:
- DHCP Server
- DFS Namespace Server
- WINS Server
- File Server (not SoFS)
- iSNS server
- Generic App/Script/Service
Determine scenarios for a SoFS versus a clustered file server
- SoFS is recommended for use by applications such as Hyper-V or SQL Server; applications that leave files open for extended periods of time.
- Clustered file server is recommended for traditional file services: user shares (aka Folder Redirection), and server applications that open and close files frequently.
- Either are equally recommended for “standard” SMB shared applications such as SMB encryption, SMB Continuous Availability, SMB Direct, or SMB Multichannel.
True or False: When adding a role to a cluster, the wizard will automatically install any necessary Windows roles or features on nodes selected to have the role.
FALSE
Necessary roles and features must be installed on the nodes prior to adding the role to the cluster.
How many cluster roles can a LUN be assigned to serve as storage for?
One, and only one.
Protocols and features supported by a clustered file server vs. a SoFS
File server:
- SMB protocol
- NFS protocol
- Distributed File Services (DFS) Replication
- Data Deduplication
- File Server Resource Manager
SoFS:
- SMB protocol
- none of the others above
- Allows an active-active file share, where all nodes can access the share simultaneously.
Prerequisites for Server to Server Storage Replica
Both the source and destination server must:
- run Datacenter Edition of Server 2016
- be domain-joined
- have the File Server and Storage Replica roles installed
- each have a volume for the data to be replicated, with the same size partition for the source and destination. It cannot be the volume the OS is installed on.
- each also have a volume for logs, each with at least 8 GB free
Additional notes:
- High performance storage, such as SSD, is recommended for log disks
- all disks must be GUID Partition Table (GPT)
When should you use Synchronous Replication for Storage Replica?
Enable synchronous replication to guarantee a zero data loss, zero recovery point objective for all replicated volumes.
Synchronous replication may lead to reduced application performance and requires 5 ms or less average round trip network latency between cluster nodes.
When should you use Asynchronous Replication for Storage Replica?
Enabling asynchronous replication is less likely to reduce application performance and allows use of higher latency, lower bandwidth, and longer range networks.
It does not guarantee zero data loss or zero recovery point objectives.