Two months ago in April (yes, it’s taken me far too long to write things up!), I managed to attend the Scottish VMUG in Glasgow, and was also privileged enough to co-present a repeat of the AWS for Beginners presentation with Alex Galbraith. Thanks guys for having us! It was a long day as I was up at 4.30am and not back home until gone midnight, but absolutely well worth it.
Chris Storrie; Intro
The day was kicked off by Chris Storrie from the Scottish VMUG team. A quick straw poll from Chris showed that there was a large number of first time attendees.
Chris introduced the leadership team, comprising of Sandy Bryce, Ian Balmer, James Cruickshank and himself, and joked that the slide listing the team was ordered in worst to best beard!
Chris went on to state that the agenda was perhaps the most community driven agenda that the Scottish VMUG have run, something I am happy to have helped with and proud to have been a part of. He then went on to encourage attendees to continue the community spirit by trying to meet someone new outside of the sessions. Chris wrapped up the session by advertising the new Slack team, scottishvmug.slack.com, and that to sign up you should drop the team an email at email@example.com. Finally, he thanked the sponsors, who were Zerto, Morpheus Data and Pure/Capito.
Joe Baguley; Keynote
After Chris, VMware’s VP and CEO for EMEA, Joe Baguley, presented the keynote. Joe’s keynote was very informative and made a lot of sense. He started by talking about industry buzzwords and how some conversations VMware have with their customers involves a desire from customers to ‘do’ digital transformation, digital strategy and digital business, without them really understanding what those terms mean and why these things are desirable. Joe said that successful transformation starts with the user, which means VMware’s discussions are becoming more and more with their customer’s customer.
From VMware’s perspective, in 2017, digital transformation is about business outcomes, in particular Business Agility and Innovation, Exceptional Mobile Experiences and Protection of Brand and Customer Trust. Those outcomes when distilled down to 4 priorities for IT end up being to Modernise Data Centres, Integrate Public Clouds, Empower Digital Workspaces and Transform Security. Joe then outlined how each of these priorities aligns with VMware’s solution set.
Joe talked about the cycle of how changes occur in business IT, and that many organisations only go through this loop every 4-5 years when they are forced to, perhaps due to an application going EOL (e.g. Exchange 2010). When they do go through this loop, it can take 18 months. Joe contrasted that to more forward thinking organisations that see the ever increasing benefits of going through this loop as quickly as possible, sometimes many times a day.
Joe moved on to discuss how back in 2006 it was predicted by 2015 everything would be running in the public cloud. That clearly hasn’t happened, and shows that the benefits of cloud don’t suit every application and that moving applications to the cloud is not straightforward. Joe discussed recent analysis which predicts cloud (both public and private) would cover 50% of all workloads by 2021, with 30% of that being public cloud, which would split roughly 50/50 between IaaS and SaaS. Clearly on prem infrastructure will be with us for sometime.
On the back of the public and private cloud growth, Joe stated that VMware’s goal is to provide a ‘RAID’ like availability for services across datacentres and in future across clouds, and that NSX and Cross Cloud Architecture are a big part of that goal.
He then wrapped up talking about what is next for VMware, which includes NFV, AI, IOT, Serverless, PaaS and Unikernals
Zerto (Sponsor Session)
Joe was followed by one of the day’s sponsors, Zerto. Zerto ran through an overview of their data protection and disaster recovery product, of which I’m sure most of the community are aware of as Zerto have been around for a while now. You can find more info on Zerto at their website.
After all of that, it was then time for a coffee break.
Adrian Hornsby; AWS Session on Integrating AWS and VMware Cloud on AWS
The main sessions tracks then begun. I headed back to the main room to listen to Adrian Hornsby who is a Technical Evangelist for AWS. He asked the room who is using AWS which showed a mixed room of AWS users and complete novices, leaning mainly towards people who haven’t used AWS before. Adrian outlined AWS’ ‘customer backwards’ approach, where features are delivered based on customer requests and demand. He attributed the growing demand from customers for hybrid cloud to one of the key drivers behind the AWS and VMware relationship. Adrian outlined the common challenges with hybrid cloud and that VMware Cloud on AWS will allow AWS and VMware to overcome these. The challenges being the following;
- Multiple virtual machine formats
- Different networks
- Operational inconsistencies
- Differing Security baselines
- Multiple monitoring and control Mechanisms
Whilst I agree with much of that sentiment, hybrid cloud has many different definitions and the AWS/VMware approach is tackling one, perhaps more focused but restricted view of hybrid cloud. VMware Cloud on AWS will allow you to run VMware workloads in the ‘cloud’ without retooling or importing virtual machines, but it still remains a traditional vSphere setup, albeit one that is now very close to an AWS datacentre (i.e. inside it). You do get Elastic DRS and on the fly cluster resizing, very cool features but ultimately its still vSphere workloads, albeit easy to consume ones, that are very close to AWS native services. And you have to manage the two separately, vCenter for vSphere, AWS console and API for AWS. VMware are also tackling hybrid cloud in another way with the very interesting Cross Cloud Architecture; being able to perform holistic management across multiple public clouds and on premises infrastructure.
Anyway, back to Adrian’s session. After the intro, he outlined some common scenarios which AWS perceive customers will look to use VMware Cloud for;
- Maintain and expand
- Consolidate and migrate
- Workload flexibility
With the intro out of the way, Adrian then did a bit of a technical drill down, outlining how the account structure would work. VMware Cloud gets its own VPC, which is managed by VMware, and then the customer needs their own standard VPC in a AWS account, which is used to for transit connectivity into the VMware Cloud VPC. There is a new type of AWS endpoint for VMware Cloud, which will be interesting to see how it works, as there is currently no transitive routing across a VPC to another VPC, unless using a third party router within the VPC. Perhaps that is what the new endpoint is doing?
Arian talked through some examples of mixing VMware VMs with AWS native services, such as using S3 and an S3 endpoint to keep this all within AWS with no internet breakout required. Other examples included being able to quickly get data into Amazon’s data warehousing service, Redshift, and being able to migrate database workloads to RDS whilst keeping the application components on VMs in VMware Cloud.
He then ran through an overview of Aurora, Redshift and some of the new services such as Polly and Rekognition. To wrap things up, Adrian performed a cool demo where he created a webpage in S3, uploaded a photo to the page which was then analysed by Rekognition after which the results were read out by Polly.
Community Round Table on vCenter Updates
The next session for me was a community round table on vCenter Updates, which was being hosted by Sandy and Chris from the VMUG leadership team, with attendance from a VCDX in the form of Rebecca Fitzhugh and representation from VMware in the form of Paul Nottard. It was a very useful session with inputs from a number of people in the room, and the topic moving around between certificates in vCenter, 6.5 upgrades, the use of management clusters and vCenter HA and its requirement for an external PSO. I also banged on about Auto Deploy for a short while :).
It was then time for lunch.
Mark Brookfield; vRealize Automation with SRM and Puppet
Following lunch the session tracks continued, I headed to Mark Brookfield’s session on automation. Mark’s session was on using vRealize Automation in conjunction with SRM and with Puppet. The session was interesting to see vRA in action, something I’m not at all familiar with. The SRM piece is useful to see as SRM traditionally has been hard to work with outside of the GUI, although Mark did caveat that you need to be careful who you give automated DR tools too, giving an example of a customer doing a full invocation rather than a test failover. The two demos Mark ran with vRA and SRM were to pick up replicated datastores so these could be filtered upon and selected when creating a new VM, and using vRA to trigger an SRM failover.
Mark also covered using Puppet in vRA. He used vRA to spin up a VM and have the Puppet agent installed with the configuration in place so that VM could then connect to a Puppet Master and continue its configuration from there.
Mark started the session with a slide joking that on average only 8% of his demos successfully work. Unfortunately for Mark, none of his demos worked in this session. He took it with good humour though, bring up the 8% slide and changing it to show 0%! Although the demos didn’t work, it was still an informative session and there was also good interaction from the audience.
James Smith; Morpheus Data (Sponsor Session)
Then it was time for a session from one of the day’s sponsors, Morpheus Data. James Smith showed one slide and then jumped straight into a live demo of the product, which is essentially a very powerful multi cloud/on-prem provisioning portal, which also supports logging and backups. I wrote about Morpheus earlier this year in my London VMUG post so I won’t repeat myself. Having not had chance to see the actual demo in London it was good to see it this time around. Definitely something to look at later in the year.
And for the final session of the day it was time for Alex and I to step up and present our own session. We had about 20-30 attendees, of which only a couple really had any prior AWS knowledge. I think the session went well and we had some good questions at the end, even though we over ran. The slide deck for the session can be viewed here.
With all the session tracks complete, we headed back up to main room for a wrap up and prize giving, then it was off to the pub for vBeers.
After vBeers, the final part of the day was the short bus/flight/train/bus home.
Slides from the days event can be found here.
Thanks for reading! Next time I hope to publish a little quicker! For some reason I have been persevering with the Windows WordPress App, which keeps crashing on right clicks and editing images. I’ve switched back to using the WordPress site directly and everything works as expected, funny that!