Home > Uncategorized > VMware vSphere Fast Track Day#5 – Lessons Learned

VMware vSphere Fast Track Day#5 – Lessons Learned


This is my fifth and last day attending VMware vSphere Fast Track training at Jean Cordier Academy in Leuven, Belgium.
You can read here day onetwothree and four.  It is time to sum up what I have learned today in no particular order.

  • No boiled coffee for me today, my stomach cannot handle it anymore!
  • My trainer, Bert De Bruijn,  is an experienced VMware trainer. He deserves 5 stars on the evaluation sheet.
  • VMware Fault Tolerant provides high availability with zero downtime. How does it work?:
    • When you turn on FT, the system creates a secondary VM (process is similar to vMotion),
    • The Record/Replay technology (vLockstep) creates log entries of all nondeterministic execution of the primary VM and store them in a circular buffer on the VMKernel,
    • VMKernel fills and flushes the log buffer asynchronously and within milliseconds sends the log entries to the log buffer of the secondary VM,
    • This communication is done through a socket on the VMKernel ethernet adapter,
    • Secondary VM has only read access to the disks, all writes are marked as completed,
  • To detect VMKernel, VM and host failure:
    • FT uses network heartbeat over the IP address used for logging,
    • The primary and secondary VMs send ping packets to the logging IP address, if one side doesn’t receive pings within about one second then that side initiate a failover,
    • To detect VM failures, the VMKernel monitors the frequency of log update from the config file and VMM, and if the log is not updated within the timeout period the VMKernel posts an action to the VMM to force it to append an entry to the log. If the VMM doesn’t respond and does not post a log entry the VMKernel initiates a failover.
  • For more info about FT, follow this link. It aggregates many links to VMware documents and links to community blog posts.
  • VMware is OEM’ing with an exclusive agreement the Neverfail technology under the name of vCenter Server Heartbeat. It provides continuous availability to the vCenter Server but at a certain cost ($$$).
  • Backup an ESXi with the vicfg-cfgbackup command which you run from a vCLI or Set-VMHostFirmware you run from PowerShell.
  • VMware VCB is an old framework phasing out.
  • vSphere introduced a new set of APIs to replace VCB called vStorage APIs for Data Protection.
  • VMware Data Recovery is a Linux appliance.
    • It is agent-less,
    • Disk-based backup and recovery tool for VMs,
    • vSphere client plug-in,
    • Leverage destination storage with RDM disks and VMDK files using deduplication technology, but also on a CIFS shares (no dedup),
    • Default username/password: root/vmw@re,
    • With VMware Hardware version 7 vDR makes a great use of Changed Block Tracking (CBT) to reduce the data to backup,
    • Read a complete description of the product here and another article regarding quiescing mechanisms here.
  • Host Profile helps with configuration management by capturing a gold config, then replicating it to hosts. It eliminates per-host configurations and maintain configuration consistency and correctness.
    • Supported only in vSphere 4.0,
    • You cannot apply a host profile to an ESX 3.5,
    • You can automate remediation by creating a scheduled tasks,
    • The host must be placed into maintenance mode,
    • Depending of the golden profile, the host you’re applying the profile to might need a reboot,
    • You can export or import a host profile (.vpf),
    • Only available with Enterprise Plus licenses AND the default built-in 60 days license -> use those 60 days to deploy your environment using great tools like Host Profile!
  • VMware Update Manager is another tool that helps keeping consistency across a data center for the hosts and VMs.
    • ESX/ESXi patches are downloaded from VMware,
    • For Windows/Linux VMs and applications, patches are downloaded from https://shavlik.com
    • Host update requires maintenance mode, and if you don’t have DRS/vMotion to evacuate VMs, you can have them shut down.
  • VMware Update Manage provides the following type of baselines:
    • Upgrade baseline – Defines what version a particular host, VM hardware, VMware Tools or virtual appliance should be,
    • Patch baseline – Defines a minimum level of updates that must be applied to a given hist or VM,
    • For each baseline you have 2 patch options:
      • Fixed – The admin manually specifies all the updates included in the baseline,
      • Dynamic – The content is determined in relation to available updates that meet the specified criteria such critical patches, security patches, …
    • You can group several baselines to form Baseline Groups,
    • Preferably attached baseline or baseline groups to container objects such as folders, clusters, datacenters,
    • To attach baselines to VMs, Templates or virtual appliances go to the VMs and Templates inventory view.
    • To attach baselines to hosts, go to Hosts and Clusters inventory view,
    • For ESX/ESXi hosts in a cluster, the remediation process is sequential,
    • Guest agent is licensed from Shavlik.com for Windows and Linux VMs. It is automatically installed the first time a patch remediation is scheduled.
  • ESX or ESXi that’s the question:
    • Why should I installed ESXi instead of ESX?,
    • Patching an ESXi means re-installation from scratch, this is the situation where Host Profile feature is a great asset,
    • You can still backup/restore an ESXi configuration, read more here,
  • ESXi partition layout looks like this:
    • Any hardware vendors diagnostic partition is retained,
    • One 4GB VFAT scratch partition for system swap,
    • One 110MB diagnostic partition for core dump. You can eventually remove this partition and use a shared storage instead (KB1004128),
    • One VMFS partition on the remaining free space. That space space should not be used in environment with multiple hosts. In the case of a USB key, that space is just not available to the host,
    • If the 4GB VFAT scratch partition is  not available, the host will use an additional 512MB of the physical memory.
  • Installing ESXi using:
  • Guided Consolidation Architecture enables you to streamline your datacenter by transforming your physical machines into VMs. This tools is available as part of the vCenter Server and many times overlooked:
    • Find – Search for and select physical server that you want to analyze (candidate for conversion to virtual),
    • Analyze – Selected physical systems are analyzed and performance data is collected,
    • Consolidate – Performance data is compared to the resource available on the ESX/ESXi hosts, selected physical system are converted and added to the vCenter Server on the recommended host

These five days attending the VMware vSphere Fast Track were awesome! I’ve learned a lot and looking forward my VCP exam. I hope you enjoy the article series and found it useful.

About these ads
Categories: Uncategorized
  1. April 3, 2010 at 18:37 | #1

    “If the scratch partition is not available, the host will use an additional 512MB of the physical memory”

    Good to know.
    Nice article serie BTW ;)

  2. Bert de Bruijn
    April 4, 2010 at 09:31 | #2

    Hi Didier, I’m honoured that you liked the training ! Still wondering what makes my sense of humor “interesting” though ;-) I expect that your VCP exam will be quite easy, given your experience. Good luck !

  3. April 6, 2010 at 19:47 | #3

    Yes VMware FT does provide high availability as described however it has some significant limitations which may be important for certain applications. The most obvious is that the guest operating systems using VMware FT are limited to a single virtual CPU . For larger applications it may not be possible to run with this limitation. For an alternative approach to fault tolerance for vSphere hardware fault tolerant servers running vSphere are a viable and easy to use option.
    (In the interest of full disclosure I work for Stratus Technologies, http://www.stratus.com. We are leaders in Hardware Fault Tolerant Servers including models supporting vSphere 4.0)

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s