Module 1.
Datacentre virtualisation (DCV) is the conversion of Hardware devices to Software resources. The hypervisor acts as a “traffic cop” allowing VMs to access underlying hardware and permits consolidation onto fewer physical hosts. Consolidating servers onto fewer physical hosts reduces the cost of power and cooling in addition to reducing capital expenditure and management costs.
Datacentre virtualisation also assists in improving availability, saleability, optimisation and system management.
DCV helps to minimise downtime from hardware failures, permits hardware maintenance and upgrades during business hours, makes backups and replication easier and permits easier scaling.
Optimisation improvements include the monitoring of performance bottlenecks and the improvement of space utilisation. Detailed performance reports are available. Storage consolidation along with storage classes depending on the VM disk IO speed requirements. Hosts can be over or under provisioned (and managed accordingly). Virtual network traffic can be prioritised.
Management improvements include the disposal of legacy hardware and centralised VM/Host/Data store management. Monitoring for regulatory bodies is also simplified.
Increased availability permits SLA improvements and simplified disaster recovery.
Scalability is leveraged from shared storage infrastructure.

Module 2.
The OS of a guest VM does not require modification to run on virtual hardware. The infrastructure can be managed from the vSphere web client. A VM can host any application, can be used for consolidation or separation, has improved disaster recovery and quick scalability. Snapshots capture the running state of a VM. A VM can have an Image Based Backup, similar to a Bare Metal backup, stored in a similar way to a snapshot for file restoration. VMs are hardware independent for restores. Hardware can be hot-added to a guest VM.
A Type 1 Hypervisor runs on bare metal (ESXi) while a Type 2 hypervisor runs on top of a traditional OS (VMware Workstation). Type 1 hypervisors offer increased stability and reduced overhead (while the stability of the hypervisor is not dependant on the underlying OS). ESXi can be run from a flash drive.
vCenter centralises management and the performance of common tasks such as creating VMs, adjusting configurations, performance monitoring and the configuration of hosts. vCenter is scalable, with components able to be installed on separate machines (VC Server, Identity Management Server, DB server, AD, Web Server, App Server).
Distributed Resource Scheduler (DRS) provides load balancing across hosts. vMotion migrates running VMs across hosts. Distributed Power Management can turn off unused hosts in periods of low utilisation and power them on again when they are needed. Storage vMotion migrates running guests across storage devices. Storage DRS is used for storage load balancing. vSphere Data Protection gives the ability to backup VMs. HA restarts a guest on another host if there is a problem on the current host. Fault Tolerance provides uninterrupted availability of a reboot is too much downtime on the guest. vSphere replication is used for remote site copying.
A VM is a large chunk of memory on a host, and exists as a file. vMotion has the ability to copy between hosts with no shared storage. Unbalanced hosts create “hotspots” and limit performance.
It is possible to define rules to be used by DRS to enforce separate hosts for certain VMs, or force them together. DPM consolidates hosts, powers down the unused host and powers back on when it is needed. DRS balances load.
When an ESXi host is configured to see both storage devices, storage vMotion can migrate disk files of running VMs. Storage DRS monitors the IO on storage devices and migrates disk files as required to maintain balanced load. Storage IO control can be used to throttle the disk IO of busy guests, to permit other guests access time to disks. VDP (vSphere Data Protection) provides image based backup with integrated data deduplication, with image based backups being the equivalent to a bare metal backup.
HA does not run on the VC, instead each host contributes to monitoring other hosts – this improves redundancy. Guest OS can have drivers installed for HA monitoring, where the guest will be restarted if the driver stops responding. Services and applications can also be monitored.
Fault Tolerance (FT) – a running VM is continuously mirrored to another host for 100% uptime. vSphere Replication offers configurable replication intervals and point-in-time counts.
Services relying on shared storage – DRS, DPM, Storage DRS, HA and Fault Tolerance. Supported shared storage types – Fibre Channel, Fibre Channel over Ethernet, iSCSI, NFS and Local. VSA (vSphere Storage Appliance) creates shared storage from local storage on an ESXi host. This permits cost savings in small environments and permits the central management of VSA clusters for SMB.
vFlash is a pool of SSDs installed in ESXi hosts which improves the disk read speed for guest VMs as well as reducing load on the SANs. VMFS datastores can be built on iSCSI, FC, FCoE and Local storage. NFS datastores are built on NFS shares. Virtual Networking is used for IP Storage, VC Management and VC communication. The Virtual Switch is then attached to a physical switch. Standard virtual switches operate at the host level. Distributed switches are managed and operate at the datacentre level.
Network features supported by Virtual Switches – VLANs, Shaping, Port Mirroring, QoS and DSCP, CDP and LLDP.
Different editions of vSphere have different feature sets. Additional VMware products: VC Operations Manager, VC Configuration Manager, VC Site Recovery Manager.
VC Operations Manager works with VC to summarise statistics onto Health, Risk and Efficiency categories. This simplifies capacity management and forecasting. Also has dynamically adjusting thresholds to prevent false positives.
VC Configuration Manager is used to assist with Change discovery and correlation, configuration analysis, correlation of changes to performance issues, regulatory compliance and can be used when hardening the environment and auditing configuration.
VC Site Recovery Manager can coordinate DR plans and DR testing, and can be configured to protect a small or large amount of machines.

Module 3.
An application that must not be interrupted by hardware failures requires FT. Minimising downtime from hardware failures requires HA. Hardware maintenance during business hours requires motion. VR replicates machines to another site. VDP tests backups and restores. VMs can be used for application isolation, and can have CPU and RAM added using the “Hot Add” feature. VC Ops is used to identify performance bottlenecks.
Thin provisioning of VM disks can improve storage efficiency. VC Operations Manager provides performance reports and trends. VM Storage Profiles are used to map VMs to storage devices / storage classes depending on IO needs. Distributed Switches can prioritise VMs with QoS. VMware can save IT expenditure, improve reaction time and agility and give automation and optimisation to infrastructure & tasks.