Uncategorized

We all enjoy going to our local VMUG meetings and hearing about new cool products but even more we love hearing from our follow peers. The community sessions are one of the strongest things of the VMware User Group.

I remember hearing about how people built their own self service portal years in advance of vCloud Director and vCAC and listening to real life experiences.

What we need is more community sessions. A lot of guys have so much to share, but not all feel confident about it. This if what #FeedForward is all about. Instead of getting Feedback (after the presentations is done, and it is possible the only time you will present it) There is much more value into getting Feedback in advance hence #FeedForward. This will make sure your presentations will improve and that you will enjoy presenting your own cool story.

Mike Laverick is the man behind the program. A new landing page was just created. http://www.vmug.com/feedforward

 People can express the intention to either:

• Becoming a Mentor and reviewing presentations for VMUG events
• Sharing a presentation at a VMUG event
• Serving on a Committee to promote FeedForward

If you want to help mentor your fellow VMUG’ers og present your self please sign up on the website.

I am personally a #FeedForward Mentor and have already helped a few guys before their first presentation.

A year has passed and VMware just announced the vExperts of 2014. A total of 754 were selected. This was my fifth straight year as a vExpert. A big thank you to VMware for showing their appreciation of the community work I have been doing.

I would like to congratulate all vExperts, but a special congratulation goes to the danish vExperts. I have spotted the following people on the list:

• Frank Brix Pedersen
• Liselotte Foverskov
• Kenneth Hansen
• Nicolai Sandager
• Mikael Korsgaard
• Gert Kjerslev
• Heino Skov
• Mads Fog Albrechtslund
• Anders Mikkelsen
• Enrico Laursen
• Rasmus Haslund

I had hoped to see Michael Ryom and Terkel Olsen on the list, they have put a lot of hard work into their blogs and I am sure we will see them soon on the list.

 

Eric Siebert posted the results of the Top VMware & Virtualization blogs yesterday. You can find the result here

This years top 10 looks like this. Once again a well deserved winner is Duncan Epping with his eminent blog Yellow Bricks.

For me personally this year was a pleasant surprise to me. I made it into number 32! That is 66 positions better than last year. Thanks for anyone who choose to vote for me. vFrank is now a top 50 virtualization blog!

 

 

It is that time of the year again. Voting on vsphere-land to find the top virtualization blogs. If you are a regular reader at vFrank.org please help by showing your support. Hopefully I will do better this year.

Vote here: http://www.surveygizmo.com/s3/1553027/Top-VMware-virtualization-blogs-2014

 

During my first two months as a Systems Engineer for PernixData I have seen a lot of similar question from customers and during POC engagements. All  good and valid questions. I felt like it would be a good thing to share it with everybody out there who are looking into PernixData FVP. 

 

Top 20 FAQ on PernixData FVP:

 

Q1: What is PernixData FVP?

A: PernixData FVP is a kernel module installed inside the ESXi hypervisor that uses local flash as a transparent cache to accelerate virtual machines storage performance.

Q2: What storage protocols does PernixData FVP support?

 

A: PernixData FVP supports block storage. FC, FCoE and iSCSI

Q3: What about NFS support?

 

A: Currently there is no support for NFS as the backend storage.

Q4: What version of VMware is supported with PernixData FVP?

A: As of writing PernixData FVP supports vSphere 5.0 and 5.1.

 

Q5: What about support for vSphere 5.5 and the new Web Client?

A: vSphere 5.5 and full web client support is coming soon. It is currently in beta. You can request access to the beta here: http://beta.pernixdata.com

 

Q6: What kind of Flash devices do you support?

A: We support every flash device that is on the VMware’s HCL list. This could be PCIe or SSD drives.

 

Q7: So PernixData FVP is a caching solution. Is it only Read cache it supports?

A: PernixData FVP supports Write Through (Read) and Write Back (Read & Write) caching.

 

Q8: Write Back caching, how do you make sure I don’t lose data in case of host or SSD failure?

A: PernixData gives you the option to choose 1 or 2 replicas partners. If you choose two your write data will be acknowledged from three SSD in three different hosts.

 

Q9: What network do you use for replication traffic. Can I select another one?

A: PernixData by default selects your vMotion vmkernel adapter. You can choose another one and put it on a dedicated vSS/vDS with dedicated physical nics.

  

Q10: Does PernixData FVP work with DRS, HA, vMotion etc.?

A: PernixData has full support for all VMware clustering features. The administrator will still be able to vMotion, use HA in case of a host failure and DRS can distribute the virtual machines.

 

Q11: Wow! Impressive, there must be some VMware features you do not support?

A: PernixData does not accelerate Raw Device Mapped disks (RDM) and PernixData does not support VMware fault tolerant protected virtual machines.

 

Q12: What VMware licenses do I need to get started with PernixData FVP?

A:  PernixData supports any VMware vSphere license. It does not matter if it is Essentials, Standard, Enterprise or Enterprise plus.

 

Q13: Is it complicated to install?

A: PernixData FVP is a very simple install. It consists of a management server running on windows, the FVP host extension installed inside the VMkernel and a plugin for the vSphere Client.

 

Q14: So I have bought two SSD’s for my hosts in my cluster. How should I configure my RAID controller? Is RAID0 good for performance and RAID1 good for security?

A: You should not configure RAID0 or RAID1. Keep it simple and direct attach the disks as JBOD. If JBOD is not possible then configure a RAID0. For best performance we prefer not to use the RAID controller at all.

 

Q15: How much flash cache do I need in each ESXi host?

A: This is not a simple question to answer. It depends on the working set of the virtual machines. As of writing we see most customers going with 200/400GB flash pr. host.

 

Q16: So I have been running PernixData for some time now, how do I know if I need more flash?

A: Inside the PernixData tab in vCenter we show you the eviction rate. If you see evicted data it means you can benefit from more flash.

 

Q17: Do I need the same flash type and size in every ESXi host?

A: PernixData is very flexible. It is a good design to use similar flash type and size to get predictable performance. But if you want to add 1TB Enterprise flash in one server and 200GB consumer flash in another we won’t stop you.

 

Q18: What kind of workloads can I accelerate with PernixData FVP?

A: Because PernixData is a clustered solution with WT and WB cache you can protect any workload. It does not matter if it is VDI or normal servers.

 

Q19: What kind of latency can expect for my virtual machines with PernixData?

A: It depends on the flash device and the workload. But microseconds and very low milliseconds range.

 

Q20: How do I get started with your software if I want to try it?

A: You can request a free trial on www.pernixdata.com and try it out for 60 days. During the trial you have access to a Systems Engineer from PernixData who will help you out with any questions you might have.

Exclusively for the virtualization community, Veeam raffles great prizes that will improve IT Pros daily working experience!

Join the raffle and be fully equipped for 2014!

Register now to get one of the following prizes:

• Tablet of your choice: Android, iPad or Surface
• Class of your choice: Microsoft Training or VMware Education Services course
• Event of your choice: TechEd or VMworld in your geography (pass, no travel)
• Home lab with HP and Netgear products: one mega server, ReadyNAS Pro, SSDs and amazing WiFi router
• Software kit: MSDN subscription, Veeam NFR licenses and VMware vSphere

Make sure to register before Dec. 24 to enter raffle!

You can check out the competition here: http://go.veeam.com/ultimate-new-years-resolution.html

A few months ago I was mentoring Rasmus Haslund on his first delivery of a VMware vSphere Fast Track Course in Iceland. During the week we had a lot of time to discuss work and what we wanted to achieve. One of our goals was to make it in to the vSphere Top 50 in 2014 on vsphere-land. To get there we knew that we needed to create quality content. Since September I have been extremely busy trying to accomplish this. Some of the articles that helped to grow the number of visitors are:

How to log in to Single Sign ON SSO in vSphere 5.5

Free VMware VCA Certification Voucher!

Understanding VMware Reservations Limits and Shares

Understanding VMware Ballooning

Take a look at my blog statistics week-by-week. It really makes a difference when you are serious.

 

I want to thank my visitors for coming to my site. Thanks for all of the comments. And a very big thank has to go out to my blog sponsors. During the last months I have been so lucky to be sponsored by

Unitrends – Who makes a very competitive backup solution for VMware. [Unitrends]

Zerto – You must check them out if you are looking into disaster recovery. They are much more than VMware SRM [Zerto]

Zitcom – A Danish hosting provider. They have some very talented guys managing their servers. I would have no problem to have my services hosted at them. [Zitcom]

Veeam – Requires no introduction. Great product for virtual machine backups that continue to grow. [Veeam]

During the last 5 years I have been working for Arrow ECS as a VMware Certified Instructor and consultant. It has been the best job I have had. I have great colleagues and I want to thank the current VMware team at Arrow Daniel Massot, Annette Jakobsen, Jakob Will Thomsen and Rasmus Haslund for great cooperation during this time. Also a big thank you to Karsten Bundgaard and Thomas Hagedorn. These guys (and one girl) are not only good colleagues but also good friends today.

Some of the key highlights my time at Arrow:

• VMware Certified Instructor Pioneer of new courses award 2010
• VMware Certified Instructor of the Quarter (Q2) 2012
• A lot of VMware certifications including VCAP-DCD, VCAP-DCA, VCAP-CID
• VMworld Speaker 2013 in San Francisco (VCM4555) and Barcelona (VCM5100)

Even though I have a great job with great colleagues it is time for me to move on to new challenges. My next challenge will be to start my own company vFrank Consulting. I will be independent from December 1st. 2013. In my new company I want to focus on what I do best, consultancy work and deliver VMware training. By being independent I will be able to go to the customers for longer time engagements. Something I have not been able to do today.

My goals for 2014 and vFrank Consulting

• Submit and defend a VCDX design
• VMworld Speaker 2014 on vC Ops
• Have fun, work a lot and learn lot 🙂

Can’t wait to get started.

During this years VMworld I was so lucky to get a signed hard copy of the vSphere Design Pocketbook. It is a vSphere design book in with all its design recommendations in “tweet factor”. The book is sponsored by Pernixdata

The good news is that if you did not go to VMworld you can have the book for free as an ebook. Just sign-up here

The book is written by top bloggers and people from the virtualization community. I highly recommend it.

Snapshots is a very strong but dangerous feature in a VMware virtual environment. You have the option to save a virtual machines state and later revert to it. 

It is a general best practice to “delete” the snapshots after a predefined time period (for instance 7 days) if you are not using them. Sometimes backup products or administrators create snapshots without cleaning them up. If you find a virtual machine with snapshots and you know they should not be there what do you do?

Simple answer: Delete All

If the current state of the virtual machine is what you want just hit the “delete all” button. It will make sure to consolidate all snapshots in the chain you are in. If you have snapshots in another chain they will just be deleted. 

Be very careful with the “Go to” button. If you hit it you will go to that state and the current state you are in will be lost. If you want to be able to go back to where you are now, then take a new snapshot before you revert to a prior.

Remember snapshots are basically delta.vmdk files on a datastore. They are like a thin provisioned disk and can grow to the same size as its parent file. That is why you need extra disk space for snapshots and why you should clean them up if you don’t need them.