VMWare cheatsheet

Running Vmware console from command line

Running vsphere client from command line

terms

• OVF (Open Virtualization Format) - from Open Virtualizatoin Format Specification, v1.10 and wikipedia - An OVF package always contains exactly one OVF descriptor (a file with extension .ovf). The OVF descriptor is an XML file which describes the packaged virtual machine; it contains the metadata for the OVF package, such as name, hardware requirements, references to the other files in the OVF package and human-readable descriptions. In addition to the OVF descriptor, the OVF package will typically contain one or more disk images, and optionally certificate files and other auxiliary files.
• VMDK (Virtual Machine Disk) - file format used for vitual applieances developed for VMWare products. Container formate for virtual hadr disk drives. Open format.
• VMotion (from http://www.vmware.com/files/pdf/VMware-VMotion-DS-EN.pdf) - enables the live migration of running virtual machines from one physical server to another with zero downtime, continuous service availability, and complete transaction integrity.
• vRAM - virtual RAM shared between images?

vSphere 5 licensing / editions / kits

• Essentials - all-in-one targetted at SMBs. Entirely self contained and cannot be used with other versions. Limited to 3 hosts, 192GB pooled vRAM.
• Essentials Plus - as above, but adds vMotion, HA, VMware Data Recovery features.
• Standard Acceleration
• Enterprise Acceleration
• Enterprise Plus Acceleration

vCenter Server

editions

• for Essentials - used with both Essentials kits
• Foundation - used for centralized mgmt of up to 3 ESXi hosts. Not sold as part of kit. ALlows purchase of individual component of vSCenter Server
• Standard - used for Standard, Enterprise, and Enterpise plus kits. No limits on restrictions of hosts that can be managed except for supported max of 1000 hosts. Includes Orchestrator and Linked Mode as additional features

install and sizing notes

Can no longer install Guided Consolidation or Converter.
Choose correct size for vCenter JVM and change after install via http://kb.vmware.com/kb/1039180.
    You can size via Administration / vCenter Server Settings in vCenter Server Settings Statistics window, 
    or via www.vmware.com/support/vsphere4/doc/vsp_4x_db_calculator_oracle.xls (oracle)
    or via www.vmware.com/support/vsphere4/doc/vsp_4x_db_calculator.xls (MS SQL)

    bundled SQL2K8R2Express can handle up to 5 hosts and 5 virtual machines in inventory.  
    max size is 10GB
    Cannot be installed as part of vCenter Server upgrade (latter needs to be manually 
    installed prior to running vCenter Server Installer

Reqs

• Appliance - deployed via OVF, supported only when deployed on ESX/ESXi 4 or 5 - needs vspher client - 7GB - 80GB diskspace req - VMDK and OVF must be downloaded
• vCenter Server on Windows - if you need SQL Server - .net 3.5 SP1 - 2 64 bit CPUs or 1 64 bit dual-core CPU, 4GB Mem, 4GB free disk space, supported db, dns entries resolving svr, no NAT from server to ESXi hosts, remote db must have 64-bit DSN, preferably svr that is member of domain, 15 char name or less, do not install on DC, static IP addr
• designate an account to run vCenter Svr services, system or user
• install account must have access to db (or admin for bundled SQL2K8R2 Express

default pw

vSPhere Client

vSphere Web Client

Reqs

vSphere Client Plug-ins

• vCenter Storage Monitoring
• vCenter Service Status
• vCenter Hardware Status
• Vmware Syslog Collector (syslog svr for windows)
• VMWare Auto Deploy - simplifies deployment and patching of ESXi hosts

Licensing

25 chars grouped 5 at a time separated by dashes
soft licenses allow ops that would result in exceding licenses cap.  hard licenses do not.
if ESX/ESXi 3.5 hosts to be managed, separate license svr req to supp theses hosts
upgrading from vSphere 4 to vSphere 5 requires new license
from vCenter Svr (via vSPhere Client) - Home / Administration options / Licenses / Management tab

vSphere/vCenter Availability reqs

• Most attractive option besides HA, vMotion, and/or DRS, might be VMWare vCenter Svr Heartbeat application
• protects app, cfg, OS, net, and hw-realted probelsm. Allows failover/failback on phy and virt platforms

ESXi

• ESXi Installable - install on compatible x86 svrs
• ESXi Embedded - install on USB / SD cards in svrs by OEM

Installing manually

• Make sure UTC is set in BIOS, not time zones
• If using Fibre Channel or iSCSI SAN storage, consider disconnecting cables to make sure not to install on wrong disk

Installing using auto-deploy

• PXE
• Make sure vSPhere Auto Deploy is installed
• install VMWare PowerCLI
• need a tftp svr, and DHCP, and vCenter Server
• TFTP zip file should be downloaded from Auto Deploy page in vCenter Svr, and extracted on directory on TFTP svr
• DHCP scope include option 66 specifying TFTP svr
• DHCP scope option 67 specify boot file undionly.kpxe.vmw-hardwired.
• set each ESXi host to be provissioned with Auto Deploy to PXE boot
• Use Image Builder to create image profile
• Write rules to assign image profile and optional host profile to ESXi hosts provisioned with Auto Deploy

additional cfg steps

NTP on ESXi host

DNS and routing

CLASS

VCenter preferable

• more mgmt capabilities,more control forcing login via AD credentials

Terms

• HA- minimal down time
• Vmotion
• SANstorage
• vMotion - move
• DRS Dist Resource Sche -uses vMotiont automatically load balance machines from host to host.
• DPM - Distributed Power Mgmt - Bring host instances up / down based on CPUandMemory...
• FT - Fault Tolerance - Primary machine, hidden secondary
• storage vMotion - move files from san to san or Local Disk to SAN
• VM Profile - where everything is set up (for each profile
• Storage DRS - canmonitor availablespace,and move things based on storage available
• vSwitch - virtual switch - virtual uplink connected to physical port - created on host
• Hot Add - can be added on the fly
• 2008 can add vCPUs and vMemory on the fly
• Memory Overcommitment - provide more memory than available (swapping?)
• VMFS - Virtual Machine File System - needed to store (LARGE) files -block sizes in 4.x 8MB - do not create on NAS (use NFS mount) create onother 
• Thin Provisioning - grow into storage requirements - NAS already do thin storaging
• Distributed Switch - create one switch, and replicate to all hosts

Thinking about taking vcp5

• Services/ Get Certified / VCP5 / ExVCP5 Exam Blueprint / VCP5 MockExam
• Test King
• Support & Downloads /Data Center / VSphere ESXI/Configuration Maximums for VMware vSphere 5.0( & documentation)
• review Basics Guide & Resource Mgmt Guide

Sizing

• old rule of thumb - #vCPU x 2 = # of logical cpus

• # Sockets * #cores/socket x 2

Configuring machine

• Creates .vmx file
• .log files kept (up to 8)
• .vmdk (small) descriptor file
• flat.vmdk (Large)- logical disk
• .NVRAM file for bios/uefi
• .vswp (Swap file for memory = amount of memory configured for windows - don't necessarily read real memory - don't want to use)
• separate from windows swap file
• .vswp (2nd) for overhead of file (resources machine uses when running (e.g. kernel?))
• snapshot files
• template file
• .vmss suspend state file
• .vvmsd snapshot files

vSphere User Interfaces

• vCLI - set of cmds run fom remote linux or windows system adnexecuted on an ESXi host, packaged as an app
• vSphere Management Assistant (vMA) - 
• platform running vCLI, vSphere SDK for PERL, VMware vSphere API
• Packaged asvirt app based onLinux
• vSphere PowerCLI - automation tool for admin vsphere env, distributed as a snap-in to Windows Powershell

Logs

ESxi overview

• VMM will automatically recognize to run better
• 1 place to manually set to force direct to physical process

• if kernel runs short on memory, kernel contacts baloon driver (installed with vmware tools

DCUI

• Cfg root password
• cfg lock down mod
• cfg mgmt network
• vswitch0 wil have vmk0
• Restart Mgmt Net
• Test MGmt net
• restore net settings
• config keyboard
• troublshooting options (enable ssh here)
• view system logs
• view support info
• Reset sys cfg

NTP/DNS Routing

Creating Virtual Machine

Virtual Machine Hardware (v8)

• keyboard
• mouse
• 1-2 flopy drive s 1 floppy controller
• 1 usb controller, upt to usb drives
• hd audio
• up to 4 ports - serial/com ports
• up to 3 ports
• / paraellel port
• 1 IDE controller , up to 4 devices
• 3d graphics
• video
• up to 1 TB RAM
• VM Chipset 1 CPU (up to 32 CPUs with virtual SMP)
• 41-4 SCSI adapters w/ 1-15 devs each
• 1-10 nics

• Change windows 2008 to LSI Logic SASsnapshots allowed

• independent disk - persistent or nonpersistent (disk is read only, sets asside delta file
• thick provisioned lazy zerioed (only zero out space used (on-demand)),thick provisioned eagered zeroed (zero outwhole image) or thin provisioined (only use when needed)

• Flexible - canfunction as either vlance (PCNET32 driver) or vmxnet (used to be work after installing vmtools)
• e1000 (w2k3 defaultedto this ) became e1000e (better performance
• vmxnet, vmxnet2, vmxnet3 (best today)availalbe only with VMware tools)
• whenver possible choose vmxnet3
• vmxnt2 (enahnced vmxnet) adapater with enhanced performance
• vmxnet3 builds on vmxnet2 adapater

• Connect to CDRom
• SCSI devices up to 60

Custom Cfg

Raw Device Mapping

installing Guest OS

Virtual Appliance

VCenter

inventory

• each datacenter has its own hosts virt machinees, templates, etc
• set up folder/file structure under to store guest installs

sections

events

Networking

VMkernel ports needed for - exam question?

vswitch

• hostcan have 4096 - 8 for host =4088
• can create256 switches per host
• may want to do this for security or redundancy purposes
• 802.1Q VLAN tagging supported
• stripped before going to virtual machines
• 802.3AD for NIC teaming supported
• network port security
• CDP

• vNetwork standard
• vnetwork distributed

• Promiscuous Mode
• Mac Address Changes - normally set by kernel - affect traffic coming towards vm guest os 
• Forged Transmits - allows outbound - affects traffic coming from vm guest OS

• only on outbound traffic not incoming

• Originating Virtual Port ID (Default)
• Source MAC Hash
• IP Addr (Src/Dest) Hash

• defaultRoute based onoriginating virtual port ID

• Link state
• LInk state and Beacon

Storage Concepts

Disk Drive Companies

Considerations

• only support RDM and virtual swap files

• NFS doesn't support Raw device Map (RDM) or ability to remote boot

VMFS

• 256G  - 1MB block size
• 512 G 2MB block size
• 1T - 4 MB block size
• 2T - 8MB block size
• block sizes broke into 64kb chuncks

• Can only upgrade to VMFS5 if all guest OSes are in VSphere 5
• Allows concurrent access storage
• cnan be dynamic expand
• 1MB block size
• subblock size 7kb
• on-disk,block-level locking

NFS

• NFS v3 over TCP Supported EXAM QUESTION
• storage shared over network at file system level

Storage Device Naming Conventions

• SCSI ID
• Canononical name - Net Addr Auth ID is unique LUN id, garaanteed to be peristent across reboots
• can also be ided wiht mpx or t10 ids
• runtime name - uses convention vmhbaN:C:T:L.  Not persisten through rebooots
• LUNs can be presented differently to different hosts

Physical Storage Considerations

• LUN sizes
• I/O BW
• Disk cache paams
• zoning and masking
• Identical LUN presentationsto each ESXi host
• Active/Active or Active/Passive arrays (# of controllers and paths are active)
• Path Selection Policy (PSP) - fixed/preferred for active active
• can be round robin (?)
• if you set this to MRU yuo can caus array thrashing
• PSP is Most recently used (MRU) or Round Robin (RR)
• Export properties for NFS datastores
• administrator has to change attributes to allow root access to exported folder

iSCSI

iSCSI initiator

• software iSCSI (iSCSI initator, TCP/IP, NIC Driver all in kernel...NIC is only HW componenet
• dependent hardware iSCSI (shares responsibilites with VMkernal) - iSCSInet cfg, NIC driver
• independent hardware iSCSI (vmware 5 no longer needed to remote boot fromindependent hw iSCSI)

Configuring

1. Cfg VMkernel port
2. enable sw adaptor
3. cfg IQN and alias
4. cfg iSCSI sw adapter properties, such as static/dynamic discovery addrs and iscsi port binding
5. cfg iSCSI security (eg.g. CHAP)

• do notuse CHAP
• DO not use CHap unless target requires
• use CHAP unless prohibed by target
• use chap

NAS/NFS

multipathing

1. cfg 1 vmkernel port
2. use adapters tattached to same phys switch to cfg NIC teaming
3. cfg NFS svr with mult ip addrs (can be on same sub)
4. mult links cfg NIC teamswith IP hash load-balancing policy

Fibre Channel SAN with ESXI

1. install fibre channel adapters in ESXi host
2. adapters are recognized by host during boot sequence

VMFS Datastore

• optimized for storing / accessing large files
• maxvol size 64TB
• NFS stores greatforstoring virt machines, but some functions no suppeorted
• use RDMs if virt machin
• performing SAN snapshotting
• clusttering to a phys mach
• large amounts of data that you do not  want to convert intoa virt disk

Increasing size

VSA - vSPhere Storage Appliacne

Clones

Templates

Virtual Machine Options

• Delay power on
• Boot into BIOS
• Retry afterfaild boot
• CPU/MMU Virtualization
• Swapfile locations - use defaultsettings (with guest's files)

Migrating Virt Machines

• cold - poweredoff
• suspended
• vMOtion - migrate virt machine that is powered on
• Storage vMotion - migrate virt machines files, while virt machine is powered onto a different datastore

• must not have connection to internal (lonly vSwitch)
• must not have conn to virt device (such as CD-ROm or flopy) with local image mounted
• must not hvae CPUaffinity cfg'ed

• 128 concurrent vMotion migrations per VMFS datastore
• at least 1Gbps network (4 concurrent migrations), 8 in 10Gbpps
• Identical named port groups connected to same physical networks
• Compatible CPUs
• CPU feature sets of both source and desthost must be compatible
• Some features can be hidde using Enhanced vMotion Compatibility (EVC) of compatibilty masks

• cont care about clock speeds cach sizes,number of cores
• do care about Mfg, family, presencs of multimedia sinstrucgtions (e.g. SSE3, SSSE3, SSE4.1)
• do care abuot Virt HW assist codes in cpu
• do care about NX/XD bit disabled/enabled
• can reset hiding or exposing in Guest settings 

• Look at Maps tab under machine
• red x - make sure vmotion enabled in settingsand both machines can see shared disk
• if it doesn't work, will get error msg indicating problem

StoragevMotion

1. initiate storage migration
2. copy data usigng VMkernel data mover or vSPhere Storage APIs - Array Integration (VAAI
3. start virt machin process
4. mirror IO calls to fileblocks that havealready been copied to virt disk on dst data store
   

Snapshots

1. freeze flat vmdk file,keep delata vmdk file
2. snapshot list  (.vmsd) keeps track of virt machines snapshots

• will only delete snapshot (even if snapthot in middle of chain)
• delete all deletes everything below 'you are here'
• no longer keeps all when deleting all from end of chain
• snapshot consolidation - method used to commit chain of snapshots to original machine when snapshot mgr shows no snapshots exist, but desta files still remain onthe datastore
• will show up on summary tab ' virtual machinedisks consolidation is needed'
• machine right-click - Snapshot/consolidate

vApp

• iscontainerfor one or more virtual machines
• group control

Remove Virtual machine

Converter Standalone

• convert hyperv to vsphere
• move from data cntr to 2nd data cntr
• windows /linux sources
• hyperV 
• 3rdparty image formats
• can export the other way
• enables
• reduce time to set up new virt machines
• migrate legacy svrs to new hw wout reinstall
• migrate across hetero hw
• ...
• put on separate desktop from vmware
• install on Windows w/ min 525MB disk space for components.
• sourcemachine reqs 350MB
• steps
• clone
• install sysprep if necessary
• install converter agent
• volume shadow copy/shapshot
• create new hsot and start copying disk image to new datastore
• hw changes
• cpu mod / serial
• mac addr
• graphc card
• ...
• sys reconfig
• can do hot clone (while OS is running).

roles and permissions

Reserved resources

• shares -low/normal/high/custom/500/1000/2000 only works if there is contention for resources
• Reservation (how muc MGHZ of processor) up to 1 core
• memory shares, reservation (MB), Limit
• reservation up to allocated memory
• Allocated memory can use swap
• reserved memory guarantees memory rest can swap
• 1Gig allocatd 512MB Reserved, can power 8 machines in 4 gig of memory
• 1Gig allocated 512 Reserv, 724 MB limit (latter can come from phys ram or swap, but above limit must come from shares or vswap or balloon activitiy)
• Resource Pool
• CPU Shares Reservation, Limit
• Memory resources - Normal

Resource Mgmt and Monitoring

memory

• minimum mem avail is overhead memory
• swap file size equals the diff between allocated and reserved mem

• mem pages that are candidates for swap to disk are targetted for compression
• decompressing acomprfessedpageinmemory is faster than reading from disk
  

Virtual SMP

• sockets * cores * 2 if hyperthreading
  

Resource Contention

• 1000 3000 1000 (2nd item will get 3/5ths of resource)
• kernel will try tokeep svcpu on same core, as NUMA nodes try to tyie to same RAM
• vmmemctl (baloon driver from vmware tools)

Resource Pool

• not contentian based- guaranteeing at the pool level
• expandable reservation might allow a rogueapplication to claim all unreservedcapacity
• expandable reservation - allows pool to grabmoreresourcethanreserved can be grabbed from chjild resource and direct parent resource pool

admission control for VM

• are resoruces available, or can they be gained or borrowed

Resrource monitoring

• Task Manager
• Perfmon
• resoruce charges in vCenter Server
• if CPU constantly high, may beCPUconstriained
• check CPU Ready graph of virtual machine - (ready is vCPU waiting for real core)
• memory issues
• check to see if balooning (under performance tab)
• look on host for lots of swapping
• check virt machine for active memory
• disk latency
• kernel latency
• kerneel command latency- avg tim spent inVMkernael per SCSI command
• high numbers (> than 2-3 ms) represent either overworked array or overworked host
• phys device command latency
• avg timefor phys device 
• > 15-20 seconds bad
• VMware performance charts - OVerview panel
• advanced performance charts (under performance tab)
• rollups conv funct betw statistic intervals(e.g. 5 minutes of past hour stats converted to1 past-day value)
• Setting Log levels (none, error, warning, info, verbose, trivia)
• save chart as graphics

Alarms

• Condition based trigger (number based)
• event based trigger (something happened)
• default alarms for everything
• to create alarm, right click on inventory obj, select Alarm/Add Alarm
• General tab - monitor for specific conditions or events
• trigger 
• econtdition trigger -virt machine currentsnapthot over 2GB
• e.g. host using 90% of total mem
• datastore hasbeen disconnected fromall hosts
• yellow and red warnings
• reporting tab
• range, (repeat when condition exceeds range) 
• repeat every ?? minutes
• actions
• email, snmp trap, run command
• EXAM question - SMTP server, back-end sender account
• Alarms stored in alarm tab
  

HA

• cluster the db
• set up multiple redundantADsvrs
• use vspher HA to protech vcenter virtualmachine
• use vcenter server heartbeat

• minimal downtime
• works with all supported guest operating systems
• workswith all supported sxi hw
• provide protection against failure ofa nESXi host

• 1st 5 hosts becomes primary (knows about all hosts, cfgs, etc.)
• participates in HA
• rest of hosts start out as secondary
• you can promote secondary to primary

• isolation response mode - 
• files are locked
• ha agent loses heartbeat
• ha agent looks at files,and doesn't see time stamp

• 1st host becomes master, restare secondary, if master fails, election for new master occurs...
• vcenter critical to functionality
• ha can restart hosts or application via vmwaretools
• applications need VMware tools andinstalledpiece from mfg
• relies on SAN/NAS/ shared storage to function
• HA is a cluster feature

1. enable HA
2. Host Montitoring sttaus
3. Admission control (power on operations that violate availability constraints (e.g. cpu or chassis running guest OS rules)
4. Admission control policy 
1. (how many host failures cluster tollerate)
2. % of cluster resources reserved as failover spare capacity
1. slot size is avg # of resources to meet requirements (used to be 256MHz and 256 MB default) HA will monitor and look at avgCPU/Memory requirements
2. reservations will make HA assume everyone needs this reservation size

• can have up to 5 datastore heartbeat locations (?)
• if host get's olatded, goes into isolation response mode(as set)
• can implement redundant heartbeat networks
• implement redundant isolation addresses

• insplit network situation,will never talk to 2nd (locally elected) master

Fault Tolerance

• hidden guest running
• does not work with all hw supported - needintel vt or amdv instruction sets
• one becomes primary, secondary hidden running onsecondary
• vLockstep technology coordinates EXAM QUESTION
• ideally 2x2 physical nics for mgmt and heartbeat and other
• any writes on primary get buffer duplicated tosecondary (which r/o link to shared storage)
• when primary goes down, secondary becomes primary, and another instance comes up as secondary
• guidelines
• make sure you have enough hosts
• disable power mgmt features running in BIOS
• Fault Tolerance reacquires only one CPU
• everything has to be done from primary.
• DRS will always puts primary and secondary on different machines

Host Profiles

• quick way to make changes to many hosts
• CPU
• memofry
• storage
• networking
• licensing
• (something else)
• force hosts to look at host profile and update
• cluster scans againsthost profile
• tab is cluster / profile compliance (CHekck Compliance Now)
• Apply profile - forces compliance
• host has to be in maintenance mode - 
• create/check/update/import answerfile

DRS - Distributed Resource Schedule

• manual - suggests recs you move
• partially automated - auto place hosts at power on,andvcenter migration recs for virt machines
• fully automated - automattically placed at power on and auto migrated
• can be conservative to aggressive (aggressive does everything, conservative does less - recommend leave in center between - default)
• aggressive doesn't always work as it by default moves most heavily used machine

• EVC turn on will exclude nonEVCed hosts out of the cluster
• all CPUsfrom single vendor (Intel or AMD)
• use IntelCPUswith core 2 micro architecture andnewer
• use amd 1st gen opteron cpus ornewer
• chost must be hw virt enabled
• hosts...

• swap files can be insame dir as virt,or in datastore specified by host
• affinity rules - can keep machines on same hosts
• anti-affiinity rules
• preferential rules aresoftly enforced nd can be violated if necessary
  

Adding Host to Cluster

• cannot create resource pool under host if added to cluster
• Grafted from is when you add host and drag resource pools
• DRS tab on cluster to look at recommendations, edit cluster properties, etc.
• any recs that can't be applied after attempting to apply, will show up in the Faults tab/button
  

Removing host from cluster

HA might no be able to fail over machines

• HA admision congtrol is disabled
• reqquired virt machines to hosts affinity rule prevents HA from fialoing over
• Sufficient aggregated resources exist, bu t they are fragmented across hosts
• HA uses DRS to try to adjust cluster by migrating virtual machines to defragment the resources

DPM (Distributed Power Mgmt)

• Operates on ESXI - hosts can be awakened from standby mode with WOL packets or IPMI based remote power on (WOL send from antoher esxi chost in cluster
• will poweroffhost if cluster load is low - considers a 40 minuteload history
• all virt machines on selected host migrated to other hosts
• if loadishight,considers 5 minute load history
• suggested that Intel SpeedStep and AMD PowerNow are powered off

Distributed Switch

benefits

Port binding

• configured at port gorouplevel in settings
• static- defined directly agsinst vm, don't give port up when powered up - Default
• Dynamic port 1st virtual machines powered on grab port
• ephemeral - ports power on as you power on virtual machines

VLANs and PVLANs

• promiscuous - can communicate together and outside
• isolated - cannot communicate together, can connect out thtrough promiscous
• community 0can talk together , but cannot go outside - must connect out through promiscous

Network I/O Control

NetFlow

Port mirroring

Storage APIs

• HW Accelerations APIs
• Array thin provisioning APIs

• plugins for vcenter todetermine capabilities of SAN
• home/storageproviders

• classify datastores based on levels of capabilit
• can have multiple vm storage profiles per machine
• matches up with host compliance to drs make complieant
  

Storage I/O Control

• Datastoresmust bemanaged by single vCenter 
• no extents

Datastore cluster

• do not mix VMFS and NFS
• can mix vmfs3 and vmfs4
• do notmix replicatd datastores
• relationships - any host to any data store cluster
• storage drs can manage using storage vMotion
• when virt machines created cloned or migrated
• you select datastore cluster, and DRS takes care of the rest taking into account default affinity and anti-affinity rules
• Migration recs are executed
• when IOPS response time is exceeded
• when space util threshold exceeded
• ....
• space util threshold checkedevery 5minutes
• IOPS laod history checked every 7 hours
• storage DRS selects datastore based on util and IOPS response...
• afffinity and anti-affinity rules kick in

• recommendations show up in storage DRS tab, and you can apply from there

• storage DRS maintenance mode, means put datatstore out o f use in order to service it.
• don't schedule during backups
• everything supported, but ESXi5(+) onlyl

Patch Management

• automated patch downloading
• begins with info only downloading
• scheduled at regualr intervals
• contacts sources for patching (e.g. hostupdate.vmware.com, 3rd party patch location
• Creation of baseline andbaseline groups
• Scanning
• inv systems scanned for baseline compliance
• remediation
• inv systems thatare not current can be auto patched
• reduces number of reboots required after VMware Tools updates

• db, update mgr svr, vcenter,...
• must be in stalled onwindows 64-bit machine
• to install, start vmwarevcnterinstaller,and click vmwaer vspheere update manager
• info needed during the instlalation
• hostname username ....
• settings
• networkconnectivity
• downloadsettings
• downloadschedule
• notification check schedule
• virtual machine settings
• ESX HostCluster Settings
• vApp

• fixed (static),Dynamic (alwayscurrent)

• all elements scan against baseline
• remediate on scheduledbasi
• can manually scan
• version 5 can do multiple hosts at one time

• at host consider disableingDPM,HA,andFT
• if patch is recalled, binary stripped out of repository, ifyou've already installed it, you need to manually uninstall

1. update mgr puts host into maint mode
2. DRS moves virt machines to aviail host
3. update mgr patches host and exist maint mode
4. DRS moves virt machines back per rule

Checking physical MAC addresses

# esxcfg-nics -l 
# esxcfg-vswitch -l
# esxcfg-vmknic -l