July 2016 archive

One of the biggest challenges of backing up your virtual infrastructure is long term data retention and storing data off-site for maximum protection. In the old days we used to send tapes offsite for storage, while that worked it was inefficient, costly with a lot of administrative overhead. Today your virtual environment is surrounded by the Internet of Things and as a result storing backups off-site has become much easier due to the rise of public cloud-based infrastructure. Further complementing this is the shift towards disk-based backup targets which provides greater flexibility, more recovery options and faster recovery, in addition disk-based replication provides a quick and easy mechanism to move backup data across data center or to public clouds.

Doing backup and replication to an off-premise public cloud has a lot of advantages including having your virtual environment and backup environment physically separated by distance, no ongoing capex or opex costs for a backup infrastructure and decreased administration. You also get the benefits of having disk-based backups and having them off-premise so you can easily recover if something happens at your primary site. This is especially beneficial to smaller companies that may not be able to afford the cost of implementing a backup infrastructure and may not have the expertise or time to manage it.

To make backing up data offsite even easier many modern backup applications such as Veeam Backup & Replication allow you to extend your backup repositories to cloud-based service providers. Veeam introduced a program called Veeam Cloud Connect that allows Service Providers to use Veeam Backup & Replication to offer cloud repository as a service and disaster recovery as a service to any Veeam customer. Veeam Cloud Connect empowers Service Providers to set up their cloud infrastructure so that tenants can send their VM data to the cloud and store it there in an easy and secure way.

Veeam Cloud Connect (VCC) has two deployment models, VCC Backup and VCC Replication which can be used based on a customers requirements. In the Backup deployment model a customer (tenant) deploys a Veeam client at their site which utilizes a cloud backup repository at the SP site to write backup data to. A Cloud repository is a storage locations in the SP cloud that store backups of tenants’ VMs. Cloud repositories can be used as primary storage locations and secondary storage locations to meet the 3-2-1 backup best practice. This model is depicted below:

The Replication deployment mode is setup a little different, since you are replicating you need to have dedicated computing, storage and network resources in the SP virtualization environment. To set up replication components, the SP configures hardware plans and subscribes tenants to one or several hardware plans. For tenants, hardware plans appear as cloud hosts. Tenants can create VM replicas on cloud hosts and fail over to VM replicas in the cloud in case of a disaster on the production site. This model is depicted below:

Of course to make this all work you need a good SP that supports Veeam Cloud Connect. ServerCentral is a managed data center services provider that provides managed services as well as hosting, co-location, cloud, and network infrastructure products for any size customer. Founded in 1999 they have helped thousands of businesses simplify their IT operations through reliable data centers & onsite support. They have data centers located all over the world including IL, VA and CA in the United States as well as in the Netherlands and Japan. They provide a wide range of managed services including the following:

• Disaster Recovery as a Service (DRaaS)
• Data Center Migration
• Backup & Replication
• Networking
• Cloud Storage
• DDoS Mitigation
• and much more!

ServerCentral is a Veeam Cloud Connect Service Provider with pricing starting as low as $50/TB a month. To learn more about ServerCentral’s VCC service check out their VCC web page and signing up is as easy as filling out this simple form to get you started on the road to cloud-based backup and recovery. Besides VCC ServerCentral also provides additional DRaaS options for virtual environments.

You can check out the following links to learn more about how Veeam Cloud Connect works and the VCC offerings that ServerCentral provides:

• Veeam Cloud Connect Backup for Service Providers (Veeam web info page)
• Welcome Veeam Cloud Connect! (Veeam blog post)
• Veeam Cloud Connect Administrator Guide (Veeam docs)
• Veeam Cloud Connect with ServerCentral (datasheet)
• Configuring Veeam Cloud Connect with ServerCentral (ServerCentral video)

VMware initially made 3 sessions on VVols publicly available right after VMworld last year on YouTube, my session was fortunate enough to have been one of those. All the non-attendees missed out on all the other great VVols sessions at VMworld 2015 until now. Since VMware recently made all VMworld 2015 sessions publicly available so anyone can watch them you can check out the ones you might have missed along with all the other great session content from VMworld last year.

Ben Meadowcroft over at VMware has been doing a series of posts on the Virtual Blocks blog called VVols Reloaded to highlight those sessions. I’ve also listed all of the VVol sessions below as well for you to view. For more info on VVols also be sure to check out my huge VVol link page and if you are going to VMworld this year read this to get a summary of all the VVol sessions that will be at the show.

STO4649 – Virtual Volumes Technical Deep Dive (Ken Werneburg, VMware – Patrick Dirks, VMware)
STO5888 – Top 10 Thing You MUST Know Before Implementing Virtual Volumes (Eric Siebert, HP)
STO5074 – Explaining Advanced Virtual Volumes Configurations (Ken Werneburg, VMware – Andy Banta, SolidFire)
STO4452 – Virtual Volumes (VVOLS) a game changer for running Tier 1 Business Critical Databases (Sudhir Balasubramanian, VMware)
STO5822 – Putting Virtual Volumes to Work — Storage Best Practices for vSphere 6 and Beyond (Howard Marks, DeepStorage)
STO6263-SPO – Veritas Technologies: How to use vSphere 6 and Virtual Volumes (VVOLs) the right way (George Winter, Symnatec)
STO6285-SPO – Beyond the Hype: The Customer Experience with Storage QoS (Ben Bolles, NexGen Storage)
STO4805 – How Virtual Volumes Will Solve Storage Challenges in the Datacenter (David Glynn, Dell Storage)
STO5522 – Virtual Volumes Technical Panel (US Version) (Ken Werneburg, VMware)
STO5522 – Virtual Volumes Technical Panel (EMEA Version) (Ken Werneburg, VMware)
STO5571 – What’s New in Virtual Volumes (Michael Haag, VMware – Juan Novella, VMware)
STO5721 – Virtual Volumes on NetApp – Ready For Prime Time! (Peter Learmonth, NetApp – Ken Werneburg, VMware)
STO6284-SPO – New Ideas for Simply Better VM Automation, Business Continuity and Data Protection with and without Virtual Volumes (Paul Morrissey, HDS)
STO6559-QT – Cloud-like Operational Efficiency with Virtual Volumes (Juan Novella, VMware)
STO6671-SPO – VMware & IBM: Optimized to Manage Oceans of Data in a Digitally-Driven World (Eric Herzog, IBM – Paul Braren, IBM – Carlos Fuente, IBM)

The VMworld Schedule Builder (US) is now available so time to go build your schedules and try and cram as many sessions as possible into all too short a time frame. Before I give you my picks, just a reminder that all the VMworld 2015 sessions are now publicly available for viewing. So now onto this year, I did a post a few weeks back that detailed the session breakdown this year. The numbers have changed a bit since then as some additional sessions were added but the large majority of the sessions are still VMware only sessions with very few (10%) independent speaker sessions. Despite who is presenting at VMworld there are still loads of great sessions and I thought I would detail my top picks this year by category. First up sessions that cover VVols, I did a post already that detailed all the VVol sessions at VMworld this year but here are my top picks:

Disclaimer: I’m more biased towards storage sessions and not really that interested in VDI or Networking

VVols sessions:

Transitioning to VVols: Partner Panel [8619] (Panel of experts on VVols myself included)

• Pete Flecha, Sr Technical Marketing Architect, VMware
• Several TBD technology partners

Virtual Volumes Technical Deep Dive [7645] (Learn from two very knowledgeable experts on VVols)

• Patrick Dirks, Sr Manager, VMware, Inc.
• Pete Flecha, Sr Technical Marketing Architect, VMware

Snapshots Suck: How VSAN and VVol fix all your operational nightmares [8159] (Snapshots change for the better in VVols, learn why)

• Pete Flecha, Sr Technical Marketing Architect, VMware
• John Nicholson, Technical Marketing Manager, VMware

Achieving Agility, Flexibility , Scalability and Performance with VMware Software Defined Storage (SDS) and Virtual Volumes for Business critical databases [7549] (Tier 1 apps on VVols, interesting to hear the performance angle)

• Sudhir Balasubramanian, Senior Solution Architect – Data Platforms, VMware
• Mohan Potheri, Sr Solution Architect, VMware

VVol and Storage Policy-Based Management ? Is It Everything They Said It Would Be? [STO9054] (Customer perspective on VVols)

• Ben Bolles, VP, Products, Pivot3
• Jeremiah Francis, Director, Information technology, Financial Advocates

Other Storage sessions:

A Day in the Life of a VSAN I/O [STO7875] (I love geeky I/O deep dive sessions)

• Duncan Epping, Chief Technologist, VMware
• John Nicholson, Senior Technical Marketing Manager, VMware

An Industry Roadmap: From storage to data management [STO7903] (Good to hear insights from a VMware Fellow)

• Christos Karamanolis, VMware Fellow – CTO of Storage and Availability, VMware

Storage for Virtual Environments – What’s new and What’s next for vSphere Storage [STO7755] (Sounds interesting)

• Sudhanshu (Suds) Jain, Product Management, vSphere Network and Storage Interconnect and Virtual Platform, VMWare
• SHASHANK RAJVANSHI, Product Line Manager, VMWare

Storage at Memory Speed and the Amazing Future of Virtual Non-Volatile Memory [INF9950] (Also sounds interesting)

• Aaron Blasius, Group Product Manager, VMware Inc.
• Rajesh Venkatasubramanian, Principal Engineer, VMware

Container Sessions:

(because I really need to learn more about containers, you do to)

vSphere Integrated Containers – Learn how you can run Docker Containers, in Production, Today! [CNA8717]

• Karthik Narayan, Sr. Product Manager, VMware Inc.

From Zero to VMware Photon Platform [CNA7741]

• Paul Gifford, Staff Systems Enginerer, National Specialist, VMware
• Adam Osterholt, Staff Systems Engineer, SDE, VMware, Inc.

Containers for the vSphere Admin [CNA7522]

• Ryan Kelly, Staff Systems Engineer, VMware

Introduction to Containers as a Service [CNA7454]

• Gary Coburn, Staff Engineer – National Specialist, VMware
• Paul Gifford, Staff Systems Enginerer, National Specialist, VMware

Other Sessions:

Evolving the vSphere API for the Modern Era [INF8255] (It’s William and Alan, nuff said)

• William Lam, Senior SDDC Integration Architect, VMware
• Alan Renouf, Product Line Manager, VMware

Tech Preview: Enhanced VM Availability Leveraging vCenter and Partner Hardware Integration [INF8020] (tech previews are always cool and this one sounds interesting)

• Brian Graf, Senior Product Manager – DRS/HA, VMware
• Maarten Wiggers, Staff-1 Engineer, VMware

vSphere Encryption Deep Dive: Technology Preview [INF8856] (Native encryption is finally coming to vSphere!)

• Mike Foley, Sr. Technical Marketing Architect, VMware
• Salil Suri, Group Product Manager, VMware

Extreme Performance Series: Monster VM Database Performance [VIRT7598] (always interesting to here about large scale db performance on vSphere)

• David Morse, Performance Engineer, VMware Inc.
• Todd Muirhead, Performance Engineer, VMware

Hot Topics in VMware Research [CTO9406] (bunch of smart guys talking about research, a geeks dream session)

• Chris Ramming, Senior Director, Research & Innovation, VMware
• David Tennenhouse, Chief Research Officer, VMware
• Michael Wei, Post Doc Researcher, VMware

Extreme Performance Series: vSphere Compute and Memory [INF8089] (deep dive on compute and memory resources)

• Seong Beom Kim, Performance Engineer, vmware

vSphere High Availability Best Practices and Tech Preview [INF8045] (another tech preview, what’s coming soon)

• Manoj Krishnan, vSphere HA Tech Lead, vmware
• Matthew Meyer, Technical Marketing, VMware, Inc.

Debunking the Myths about Virtualizing High Performance Computing [CTO8120] (HPC on vSphere is challenging and interesting)

• Josh Simons, Office of the CTO, VMware, Inc.
• Na Zhang, Office of the CTO, VMware

vSphere Client Roadmap: Host Client, HTML5 Client, and Web Client [INF8172] (everyone’s favorite topic, sure it will turn into a gripe session)

• Dennis Lu, Product Manager, VMware
• Radostin Tsanev, Senior R&D Manager, VMware

Zero Downtime, 20K+ VMware vSphere 6 Upgrade [INF8374] (vSphere at large scale, interesting insights from the trenches)

• Brad Calvert, Consulting Systems Engineer, HCA, Inc.
• Lee Cooper, Consulting Platform Engineer, HCA
• Jeff Guillote, Technical Account Manager, VMware, Inc.
• Joshua Surre, AVP of Infrastructure and Operations, HCA, Inc.

Sponsors of vSphere-land sessions:

Veeam Availability Suite 9.5 Deep Dive [SDDC9461-SPO] (Anton & Doug, must-see tv)

• Anton Gostev, VP, Product Management, Veeam Software
• Doug Hazelman, VP Product Strategy, Veeam Software

1,000,000 IOPS, 0 headaches [STP9439] (learn about cool new VAIO technology)

• Sheryl Koenigsberg, Senior Director of Marketing, Infinio

Evolve Your BC/DR Technology [STP9438] (cause BC/DR is always a challenge)

• Zerto

Finally a shameless plug for my session:

Containers & VVols – a technical deep dive on new technologies that revolutionize storage for vSphere [STO9617-SPO]

• Eric Siebert, Solutions Manager, HP
• Charles Tierney, Manager, VMware Alliance, HPE

There is a Denver VMUG meeting scheduled this week on Thursday July 14th from 11:30am – 4:30pm, if you are interested in going go register here. There are a couple changes happening with Denver VMUG’s that I want to point out.

The first is the location, free locations are always preferred to keep the cost of the events down. Many of the Denver events over the years have been held in Louisville at the CableLabs facility. While this location was a bit out of the way for people coming from areas like DTC, it was a free location and had good logistics to handle the meetings. The reason we kept using this location was because it was the work location of Kevin Divine, a long time VMUG leader in Denver. Kevin was looking for a change and recently moved away to a little town in Minnesota so this facility isn’t available to use anymore. Without too many other free facility options the meeting this week and I suspect many of them going forward is now at a rented space (DoubleTree Hotel in Thornton). If there are any VMUG members or companies in Denver that have a space available to accommodate up to 100 people and are willing to host a periodic meeting please let the Denver VMUG leaders now.

The next change is the Denver VMUG & Colorado Public Sector (SLED) VMUG groups have merged. There have always been two separate groups as far as I’ve known with separate leaders and meetings. I’ve attended and spoke at one of the SLED meetings years ago and it was pretty small group back then, bringing these together makes sense to get better sponsor ROI, more attendees and easier logistics. The combined VMUG leadership for Denver is now: Jason Glumac – State of Colorado, Scott Seifert – IBM and Michael Forner – RTD.

Finally you may notice from the agenda of this meeting that it is longer, has more than one sponsor and has a customer speaker which you don’t see happen to often at VMUG’s. This upcoming meeting has 2 sponsors (Veeam & Nimble), since they have to rent space this covers that as well as lunch for attendees. It also makes the meeting longer, prior meeting used to end around 2 or 3pm and this one goes to 4:30pm. The full meeting agenda is below, so if if you’re in the Denver area go register and hope to see you there.

Meeting Agenda

• 11:30 a.m. Lunch
• 12:00 p.m. Veeam Presentation
• 1:00 p.m. VMware VTA Presentation
• 1:30 p.m. Nimble Storage Presentation
• 2:30 p.m. Customer Presentation
• 3:00 p.m. VMware Presentation

Infinio just released version 3.0 of their Accelerator product which uses server-side host caching to accelerate storage operations. This is a milestone release for Infinio as they are now leveraging the vSphere APIs for I/O Filtering (VAIO) which were introduced in vSphere 6. VAIO is to storage I/O much as the VMsafe APIs introduced years ago was to host networking traffic as it allows 3rd party products to reside directly inline with storage I/O via vmKernel interfaces instead of trying to intercept I/O through more bolt-on external interfaces. Infinio Accelerator 3.0 also has some other great new features but let’s first start with a deep dive on VAIO so you can better understand why this is a big deal.

The vSphere APIs for I/O Filtering were announced at VMworld in 2014 and later quietly introduced in vSphere 6 Update 1. I say quietly as you probably didn’t hear a lot of noise from VMware on this as it’s more an under the covers enabler for 3rd party vendors and not a flashy new vSphere feature. Don’t take that quiet introduction though as an indicator of how important this new feature is, it’s a big deal and a powerful enabler for any product that interacts with storage I/O in vSphere.

To give you a better understanding of what VAIO is all about I’m going to summarize a post that VMware did last year that does a good job of explaining it. As I mentioned VAIO is not really a feature, it’s an API framework built into vSphere similar to other storage APIs such as VAAI, VASA, VADP, etc. that allows 3rd party applications to interact with the storage I/O stream at the VM level in a certified and integrated manner. Prior to VAIO vendors had to get creative with how they tapped into storage I/O by doing things like sitting inline via a virtual appliance through storage I/O traffic. Because it is integrated into vSphere it can also be managed and applied via the vSphere Storage Policy Based Management (SPBM) engine that is used with VSAN and VVols.

The benefits of this are it allows for much more efficient interaction with I/O, it also simplifies and standardizes how vendors interact with I/O and it is easier to manage overall. The real advantage of VAIO is for applications that interact with I/O close to the host and it’s VM’s which includes uses cases such as host based caching and replication applications that need to work as close as possible to the source of the I/O. With applications and hardware that operate near the end of the I/O stream their is less of an advantage and not many use cases, this would essentially include SAN and NAS devices as they are much more distant from where the filtering occurs.

Some additional use cases would be to any type of security application, such as malware scanners that need to scan I/O in real time. In addition an application that needs to encrypt data would be another good use case for VAIO. Essentially think of it like this, any application that needs to stop and look at each I/O as it leaves or enters a VM and then do something with that I/O (encrypt, replicate, scan, cache) is a good use case for VAIO. It’s important to note though that with the first release of VAIO (ESXi 6.0 U1) only caching and replication use cases are officially supported. VMware will probably certify other use cases in future vSphere releases.

So let’s now take a closer look at where the I/O filtering occurs with VAIO. So normally storage I/O initiates at the VM’s virtual SCSI device (User World) and then makes it way through the VMkernel before heading onto the physical I/O adapter and to the physical storage device. With VAIO the filtering is done close to the VM in the User World with the rest of the VAIO framework residing in the VMkernel as shown in the below figure, on the left is the normal I/O path without VAIO and on the right is with VAIO:

When an I/O goes through the filter there are several actions that an application can take on each I/O, such as fail, pass, complete or defer it. The action taken will depend on the application’s use case, a replication application may defer I/O to another device, a caching application may already have a read request cached so it would complete the request instead of sending it on to the storage device.

There are two classes of filters right now (caching, replication) and there can be more than one VAIO filter active simultaneously. Filters are executed in class order so an I/O may hit a replication filter first before it hits a cache filter, once an I/O has made it through all the filters it continues on to it’s destination be it to the VM or to a storage device. The below diagram from VMware illustrates the overall VAIO architecture and I/O path:

VAIO is a perfect match for applications like Infinio Accelerator that are host based and need to see every I/O to perform caching to improve performance. The important role that Infinio Accelerator plays is to prevent I/O from having to travel to a destination storage device which takes time to complete especially if it’s a remote storage device such as a SAN or NAS. A single I/O can take 5-10ms to make that journey or even longer if the device is very busy so to be able to cancel that journey and have the I/O instead take a shortcut from local cache can greatly improve performance. With VAIO the job that Infinio Accelerator performs is much easier as everything is integrated into the hypervisor and I/O can be filtered faster as the filtering occurs even closer to the VM.

The difference that VAIO makes is dramatic, Infinio has performed testing and was able to achieve 1,000,000 IOPS with 20GB/sec throughput and 80 μs response time in their test environment with Infinio Accelerator using VAIO integration. Infinio Accelerator can improve overall storage performance and has some key use cases with I/O intensive workloads such as enterprise database applications and virtual desktops (VDI).

In addition to VAIO support, Infinio Accelerator 3.0 also introduces support for SSD & Flash devices which add an additional caching tier for colder cached data. Infinio Accelerator’s strength has always been the use of lightning fast host RAM as a caching tier, support for SSD & Flash devices allows them to extend that caching even further for even greater efficiency by moving colder data to persistent storage rather than expiring it. Infinio Accelerator supports the latest vSphere release (6.0 U2), is certified through the VMware Ready program and supports any storage that is supported by vSphere including SAN, NAS, DAS, VVols & vSAN. Here’s a screenshot of Infinio Accelerator in action:

If you haven’t seen what Infinio can do for you I encourage you to give them a try, they offer a free fully functional 30-day trial that installs quickly without any disruption to your current environment.  They also have a recorded product demo that you can see the product in action. Below are some links for more information on both VAIO and Infinio Accelerator.

Infinio Accelerator links

• Running vSphere 6? Why you should care about VAIO – blog post
• Infinio Accelerator 3.0 brings exceptional performance to VMware-based environments – press release
• VAIO Primer: Part I “What is VAIO?” – blog post
• CEO Reflections on our Latest Release: The more things change, the more they stay the same – blog post
• Infinio Accelerator – datasheet
• Infinio Accelerator Product Overview – white paper

VAIO links

• vSphere APIs for IO Filtering – Virtual Blocks blog post
• vSphere APIs for I/O Filtering (VAIO) Program – VMware Developer Center
• vSphere 6.0 Storage – documentation

Once again this year I commissioned a graphic designer to create a counterpart to the physical coin that the Top 50 vBlogs will receive courtesy of VMTurbo. The graphic can be used by any blogger that made the Top 50 and wants to display that accomplishment on their website. Last year I wasn’t that happy with the design, I leverage Fiverr which is a huge community of freelance designers and the one I picked last year wasn’t very creative. I was trying to replicate the look of the physical coin in a design that has a metallic look to it and the designer last year didn’t seem to be able to pull that off. This year I spent more time hunting down a better designer and the one I picked got the design spot on the first try.

So below are the finished virtual coins:

You can download the hi-res images here and re-size them to whatever works for your blog:

Top 10 Gold Coin

Top 25 Silver Coin

Top 50 Copper Coin

So the voting has ended, the results have been tabulated and here they are. There were 83 new blogs on the ballot this year, 12 new blogs made the Top 50 and and 7 new blogs made it into the Top 25. One blog fell out of the top 10 this year and the competition for the #1 spot was fierce. This year there was over 1600 votes compared to around 2200 last year. You can read more stats about this years voting here. Voters were asked to pick their top 12 favorite blogs and them rank them from 1 to 12. The votes are weighted so a #1 vote is worth 12 points, a #2 vote is worth 11 points all the way down to a #12 vote being worth 1 point. The total points for each blog were added up to determine the results.

If you missed the live results show be sure and watch the replay of the special Google Hangout results show that we recorded with John Troyer and Eric Wright from Turbonomic as we count up the Top 25 results with lots of color commentary. The vLaunchpad and Planet vSphere-land will be updated soon to reflect the new voting results. Thank you everyone who voted and congratulations to the bloggers that made the Top 25. With so many bloggers out there its a tough scene but I seriously encourage you all to keep at it, the longer you stick with it, the more people notice and will reward you with their vote. You guys are all winners, I know how hard it can be to find the time to blog but do know that your efforts are appreciated and your unselfish dedication makes a difference to a great many of people.

This year any blogger that made the Top 50 will get a special 2″ commemorative coin courtesy of Turbonomic, I’ll have a form where you can enter your shipping details up in a few days, those of you that will be at the Indy VMUG or VMworld let me know and I will deliver it to you there. Bloggers who make the Top 10 will get a gold coin, 11-25 a silver coin and 26-50 a copper coin. I’ll also be coming up with a new graphic that you can display on your website if you made the Top 50.

Next year I’m looking to change this process dramatically and make the public voting be a part of an overall bigger method of determining the top bloggers, more on that later. I will also be starting a new vBlogger Spotlight series soon that is geared towards highlighting some of the lesser known bloggers outside of the top 25, if you are interested in being one of those bloggers let me know.

Special thanks to Turbonomic for sponsoring this year and making the giveaways possible. 

Here are the overall voting results…

And here are the Category voting results…