First remember that you are on your feet for almost the entire time you’re NOT in a session and in some cases a bit when you’re in there. The continual mantra that comfortable shoes need to be your first item to pack is totally true.
There are definitely other tips to give which I’ll endeavor to share here:
- Continually verify your session list(Its subject to change)
- Verify you are in the right place and where you’re going next. This is important when rooms are assigned to the schedule. Its a good idea to grab the sessions you want, but when the rooms are assigned, make sure you don’t have to run across campus to get into a session. Aka, Dont’ be like me
- Talk to people! I’m going to bring this up almost every day that I can. Its important to recognize that you, YES YOU, are a part of this community. If you’re at VMworld you’re a part of US, and we’re glad to have you! So say hi!
- Baggage! A lot of people have mental baggage but I’m talking actual luggage here. Check-in/Check-out at your hotel can be a pain. Utilize the bell-hop, or concierge, or whatever they call it at the hotel to hold your bags if the room isn’t ready, or hold them if you checkout on the last day of VMworld, and your flight is later that night.
- Another word about Sessions. If you see a session you would rather be in, than the one you’re walking to. Go do it! Sure, you’re signed up for another session, but that doesn’t mean you’ll be shot for not showing up. One thing though, if you do this, remove the session from your schedule, as you will still get a survey to fill out. One way or the other, seats are released 10 minutes prior.
- TWITTER! This little app has changed my life. Take pictures, Use Hashtags, follow people. This is a great resource for everyday growth, and importantly for the community. As I said previously, you’re already part of the community, so say Hi!
Ok imma hop off the soapbox.. Back to sessions…
Zero to App: Building Your First Application in Kubernetes [CODE1804U]
Bill Shetti @shetti @cloudjournei.io http://www.cloudjourney.io
- In 30 min we will…
- VMware Tanzu… the “Obvious answer”
- Building a cluster in PKS(Usable for the next couple months)
- The app… Node.js app
- App deployment is already built
- Has index.jas
- Has a dockerfile to show the build of the dockerfile
- docker buld -t simpleapp was the command he ran and “-t” names the app
- Tag the app using “docker tag simpleapp: *TAGINfo*)
- Pushing it to google cloud register with “gcloud push”
- Now we build the docker image, and pushed it to a container registry
- Log into the kubernetes cluster
- verify by using “kubectl –version” or “kubectl get pods”
- If the pods aren’t coming up, you can always “kubectl describe *PODNAME*”
- “vke cluster show bob” shows the information of the kubernetes cluster in vke(pks)
- PKS brings up a load balancer for you, so in Kubernetes you don’t have to run the node:LoadBalancer but you would run it as NodePort
Why should I use Virtual Volumes? A technical review. [HBI3416BUS]
Cody Hosterman @codyhosterman
- vVols are how pure storage was designed, and others.
- with VMFS you are not protecting applications, you are protecting data stores
- how VMDL restores VMs
- as vms SCALE, Challenges arise.
- what are vVols? Granular storage for VMs.
- vVols – Every virtual machine has individual volumes on the array
- every VM gets a config vVol – 4GB – holds the configuration information of the VM
- Data vVol – for every virtual disk added to the virtual machine there is a new data vVol
- Swap vVol – is for the vm swap file. sized according to the VM memory
- memory vVol – if the snapshot is taken with the “snapshot memory” option.
- What is a Data vVol?
- A data vVol is essentially a RDM. Just a block volume with whatever file system your guest puts on it. There is no VMFS on them, there is no VMDK encapsulating them. A direct volume. A well-orchestrated one!
- VMFS Virtual disks are not first class citizens, vVols make virtual Disks first class Citizens. It increases control of your storage.
- on the array AND in VMware
- What about physical to virtual?
- Move data to Cloud Native
- The Data Plane
- how does the host see a vVol? vVols are not connected to a host in the traditional way(Not enough SCSI slots)
- Protocol Endpoint
- The Management Plane
- Storage capabilities
- VMware-Created Snapshots
- Traditionally, when you took a VMware snapshot of a VM a delta file was created, this hurt performance
- Now when you take a VMware snapshot the array does the copy. no performance penalty with FlashArray volume copy!
- vVols is Space Efficient, they are direct RDMs so you can get a lot of space back by going to vVols.
- Virtual Machine Restore
- when a VM is deleted or a virtual disk how do you recover? Recovery includes loss of RTO and RPO
- With vVols you have 24 hours to instantly recover and deleted disk or VM data on the Array itself.
- VM Cloning – 100% Offloaded
- when cloning a VM, or deploying from a template –the process is a full volume copy!
- One API call to the FlashArray = Instant Cloning
- 50GB virtual Machine? Clone time is 5 seconds!
- More Insight and More knowledge
- Ransomware encrypts the data on a VM, With vVols if you see the data reduction go to 5>1 or 1>1 someone encrypted the data, you can respond quickly
- Encryption and Reduction
- Pure now encrypts the OS with a special key.
- vVols and replication
- there is a concept of a source replication group and a target group. VMs are put into a source and fail over is run against the target.
- PowerCLI Cmdlets
- Getting there…
- Flash-array vSphere client plugin!
- Migration to vVols – Storage vMotion – Online Migration!
- More info!
- Flash-array vSphere client plugin!
I blinked and was 3 slides behind. Cody goes fast and brings so many great points that I just cant keep up. Watch the video. Man this guy is good.
Is IT Ready to Enable Success in Your Multi-Cloud Environment? [MCO1687BU]
Jad El-Zein @virtualjad Keven Lees
- IT Mindset shift IS REQUIRED
- Clouds are the new silos
- The goal is blended teams, looking at Service Oriented approach
- Processes? The good and the bad..
- Governance or lack thereof is one of the biggest things with Shadow IT, and its needed to maintain the workloads and stay within policies.
- Security and governance
- Multi-cloud consumption
- Aligning with Enterprise need
- VMware Cloud Assembly – Build, Deploy, and iterate on applications
- Setup your cloud accounts, and zones
- configure your projects(business groups in vRA)
- design your blueprints(IaC in CAS)
- Cloud Assembly – Declarative provisioning and life cycle
- IaC choice of UI or Yaml Editor
- VMware Service Broker
- Infra and App Pipelines
- Content Marketplace
- Multi-Cloud Extensibility
- ABX on-prem with Powershell Support
- Multi-cloud Operations
- Self-Driving Operations
- Intelligent Monitoring and root Cause Analysis(VROPS 8.0)
- Apps-ops and wavefront are free trials, and cloud health is a trial or POC for testing
Kubernetes and Cloud Automation Services: A Primer [HBO2769BU]
CODE DE ARKLAND AND CHRIS McCLANAHAN The OG
— Nathan Bennett (@vNerdyNate) August 28, 2019
- vRealize Automation Cloud
- Kubernetes in PKS
- Kubernetes Profiles
- Tag the profiles needed within a blueprint as a constraint for kubernetes within PKS
- Use tags within the whole platform to define how things provision, Size, region, etc. the tags are everywhere.
- PKS on the Design Canvas
- Define the whole kubernetes cluster with an easy drag-drop blueprint
- On-boarding Existing and External Clusters
- If Kubernetes is already built you can bring that in so that you can manage kubernetes cluster within the application
- Master would need to be exposed publicly to be able to manage an already built Kubernetes cluster.
- Kubernetes in the Marketplace
- Currently you can do a vsphere cloud.init template that has Kubernetes on ubuntu 16.
- Don’t need PKS to do Kubernetes things
- Cloud proxy to deploy on-prem for CAS(Maybe vRA 8 Cloud?)
- Kubernetes Zones, allow you to bring into your PKS endpoint for Kubernetes
- You can t-shirt size your deployments
- You can setup the deployments and allow “Projects” to deploy the needed deployments
- You can Gui’fy namespaces within Kubernetes, this allows you to hide namespaces from other users.
- You deploy apps in this under Code Stream.
- Setting up the pipeline in Code Stream is paramount as anything as Code is needed for specific run times. This includes the app, load balancer, and everything that needs to run for the app.
- Code Stream works with Gitlab, hub, etc. and when you push a commit to your repo, Code Stream will see it and push that through Code Stream to perform the build and run.
How to not go broke with CI/CD deployments [CODE3452U]
Tim Davis @vTimd
I got to his talk late(*bladder, and sessions were too stacked .. See pointers above*), but wanted to share his talk as its been shared on podcasts and other places.
Deep Dive – Kubernetes Resource Management for vSphere Admins [KUB1208BU]
Michael Gasch and Pranshu Jain
- Putting Kubernetes on vSphere.
- Not covering Tanzu or project pacific basically just vSphere
- Its a good idea to put the kubernetes nodes in Ha as they are not Ha setup.
- Kubernetes 101
- Pod is the atomic element of Kubernetes
- Resource management is the allocation of resources from resource provider to resource consumers
- https://k8s.af – to help with postmortems and outages, 9 out of 10 related to resource management.
- But why should I care?(other than most outages are from resources)
- Best Practices doesn’t always transfer to best practices from a higher level
- Specifying the pod specialization and resources
- Resource management in the life of a POD
- Important to utilize namespaces and then refer to those tenants with what resources are available to the tenant
- By setting the validation in the Yaml with the namespaces it allows us to know the quota of the VMs and we don’t over commit or under commit our nodes.
- Best Practice for Kubernetes #1
- vSphere HA is good for Kubernetes clusters.
- Define an admission policy for HA that will help understand the normal group and the critical(high cpu etc) group
- Finally DRS using anti-affinity VM monitoring and host issues
- Placement with Kube-Scheduler
- Assisting kube-scheduler how???
- Setting resource limits on the linux kernel VMs is critical to maintaining the cluster. So each deployment in yaml as declarative or an imperative command should have the cpu/ram statements in order to change that information
Very dense talk but well worth it. Basically run on VMs, Enable HA, Use resource limitation switches, and watch your clusters.
Prove Your Skills and Validate Your Value with VMware Certification [EDU3684U]
I know this is a more personal session, but its worth mentioning that we all need to “prove our worth” too many times. Its worth having a sign or symbol to raise when someone challenges you and wont listen. We all have those people who will argue and argue. Also as we move into higher and higher positions, this is a needed step in our growth.
Karl Childs @karlchilds
- There are quantifiable results that show that individuals get benefits other than employer benefits
- VCDX – FREE POLO
- VMware Certification Tracks
- Good analogy is boulders across a river and you’re trying to get across the river by those boulders
- What if you miss the exam? or how do you prevent that?
- Run a Skills Gap-Analysis
- Download an exam-prep-guide
Things I learned about the Certificate process:
- Foundations exam is taken online, and is open book. Still have to pay for it.
- The class can be any class. Whether it be for vRA, NSX etc. But it is a REQUIREMENT for the VCP. You can pass all the exams but you will get nothing till you do the class training.
- There are lower cost options for the class part. You don’t have to do a 4500$ spend just to get your cert. This is part of the whole #vCommunity mill, but there are those that have done this part much cheaper. Not saying the actual courses aren’t worth it, they definitely are. There are just other options. If you’re a part of VMUG ask your Leader, I’m sure they know.
- I’m going to take my vRA VCP tomorrow and take advantage of the 50% off. However, with VMworld fest happening tonight and all the other things, I wont study at all before its there. But it’ll be good to know whats out there, and what the last couple years have taught me.
Taking a huge line of buses to the party was insane. I definitely didn’t want to roll with that. Thankfully a friend shelled out on an uber and we got there while Billy Idol was rockin’. Yes Billy F’ng Idol…
I have to admit he had to rely on the audience for some parts, but the dude is still killin’ it at 60 #LifeGoals. He’s still the drunk rocker persona on the stage, and he’s still crass, and swearing… but we love him anyway cause its rock n’ roll baby.
Next on stage was One Direction…… I mean, One Republic… Man I’m old.. #InsideJoke
The first song I heard of One Republic was “Secrets” which was the end credits song after “The Magicians Apprentice” movie with Nick Cage… Yup, That movie. Regardless that song holds a specific feeling in my life and whenever I hear it I’m transported to that place. However, my highlight of the night was when we got to hear “Halo” which was written by One Republic front man Ryan Tedder, and Beyonce performed and made popular. Hearing it from a male voice like his was incredibly impressive.
I met William Lam! This guy has been a huge part of the community and has always been a force for good to the groups. Thanks man for all you do!
— Nathan Bennett (@vNerdyNate) August 29, 2019
Also PEDRO ARROW, also known as Pete Fletcha, of Virtually Speaking Podcast!
— Nathan Bennett (@vNerdyNate) August 28, 2019
FINALLY met an amazing tech guru, and another leader in the community wether he likes it or not the one and only Jad Al-Zein. His blogs and videos on vRA 7.x saved my hide multiple times. Great guy to talk to, and hes SO EASY to talk to as well!!
— Nathan Bennett (@vNerdyNate) August 28, 2019
Now for bed, lets hope the test tomorrow will come out better than I expect, would be nice to pass, but I’m going to try not to hope for it a ton… even though I know it’ll be hard to sleep knowing I’ll be doing that in the morning.