Results / Findings Workshop Review 2023-06-12

[wombelix]

Hey everyone,

I did a late night review as asked by to help you get this thing over the finish line, I see three blocker that need to be fixed prior approval in the RKE2 Cluster deployment and EKS Cluster Import area, the rest is nice if it's approved but due to the time concerns for the first event, I'm fine if that get addressed after approval.

Need to be fixed before approval, but doesn't require another review, as soon it's confirmed as fixed, I can approve the Workshop:

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/30-creating-a-cluster/32-creating-rke2-cluster#cluster-information:

• [x] It would be more user friendly to provide the values that need to be changed / entered also in text form and not only as part of the screenshot. Also what I'm supposed to choose as VPC? Or cluster Name? It's mandatory but not explained.

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/30-creating-a-cluster/32-creating-rke2-cluster#cluster-information:

• [x] From here I can see that my cluster should be named "rke2-cluster", and instead of "pool1" one step earlier it was supposed to be "cp-nodes" or "wk-nodes", maybe the VPC is in both cases then mgmt-private-1? Which I don't have, it's named "rke2-eks-cluster-workshop-rancher-private-1", those cluster creation steps are very unclear.

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/40-importing-a-cluster/41-importing-the-cluster#importing-the-cluster

• [x] Same here, steps burried in the screenshots, not really clear what to do, that could use some improvement and also all relevant values in text form in addition to the pictures. Caution: region is again us-west-2, even with a set default region in the credentials, should be clearly outlined and maybe even a bug created for the Rancher UI, doesn't seem to be right, I would expect the region associated as default region with the credentials is used here.

Not a huge thing, those findings can be addressed at a later point as well, I will approve either way:

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/21-exploring-the-cluster#step-2:

• [x] nitpick, in case of an AWS event i don't have to navigate to cloudformation in the aws account, i can just open the event dashboard. Also the name of the stack, in case of an AWS event, is "rke2-eks-cluster-workshop" and not "rke2-eks-cluster".

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/21-exploring-the-cluster#rancher-multi-cluster-manager

• [x] #41

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/21-exploring-the-cluster#rancher-multi-cluster-manager

• [x] End of the page between the last two images, it looks like that there is content missing? Last sentence says "Let's take a look at some additional ways to deploy applications below:", so I would expect more than just those two screenshots? But not a major issue for the workshop flow, it's the introduction section where it's mainly about looking around and not actual doing stuff, so fine for now.

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/22-rancher-applications#step-5:

• [x] nitpick, just a regular click on the Longhorn box will already open a new browser tag, no need for a right click

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/22-rancher-applications#step-3:

• [x] nitpick, your anchors need unique names to work, you have on this (and other sides) multiple times "step X", so this link should actually go to the neuvector section, but goes up to longhorn instead. I's not pretty but instead of just "Step 1" maybe use "Longhorn - Step 1" and "NeuVector - Step 1" to make them unique per page?

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/22-rancher-applications#step-1:

• [ ] NeuVector, Step 1, nitpick, wouldn't it be easier to let people use the "Cluster Tools" link in the left menu to deploy NeuVector to ensure the Rancher integrated version is used?

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/22-rancher-applications#step-3:

• [x] NeuVector, Step 3, maybe too much / out of scope but from a best practice POV and because Longhorn was also already installed, wouldn't it be good to configure a presistant storage for NeuVector and explain that otherwise the config only sit in memory and will be lost during a re-deployment or restart of the neuvector cluster?

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/22-rancher-applications#step-5:

• [x] NeuVector, Step 5, nitpick, regular left click is enough to open in new browser window, similar as with longhorn earlier.

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/23-kubernetes-applications#step-2:

• [x] Maybe worth to explain what the differences of fleet-default and fleet-local namespaces are and why I have to change it?

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/20-exploring-the-cluster/23-kubernetes-applications#step-7: https://github.com/zackbradys/rancher-customer-demos/blob/b8defe79d907aa1de8d594ce880437b3a53d89ad/static/fleet/fleet-local.yaml#L14

• [x] Nice, I like that you added it to the navigation and demostrate how this integration works. My suggestion: Use it to explain the audience a little more about this feature, give them an Idea how they could integrate own apps into the Rancher UI and where to find more information about it.

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/30-creating-a-cluster/31-creating-eks-cluster#upload-the-required-files:

• [ ] OK that's a little confusing, the two steps before failing with access denied, even in an aws event (see below section with the blockers) but then as next step, without mentioning that I only have to do that in my own account, I get the instructions to upload those files. I suggestion to split it up, one time "cloud9 setup in an aws event" and then "cloud9 setup in your own account".

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/30-creating-a-cluster/31-creating-eks-cluster#modify-cloud9-instance-iam-role

• [ ] I have the feeling that this step is super confusing for most of the people, why it's necessary to disable temp credentials and assigning a differente instance profile should probably explained a little better to provide some context. Main reason for someone to do such a workshop is to learn something, this counts as well :)

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/30-creating-a-cluster/31-creating-eks-cluster#step-4:-create-the-cluster

• [ ] Maybe the "kubectl get nodes" command also as code view so that people can copy/paste it and not only seeing it on the picture? And btw, I hope you have a plan what to do in the Workshop at this point to bridge the wait time "Note: Cluster creation can take up to 25 minutes"

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/30-creating-a-cluster/32-creating-rke2-cluster#step-3:

• [x] Wrong picture? shows the cloudformation stack output with a mark around the rancher URL, but it's about the credentials at this stage.

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/30-creating-a-cluster/32-creating-rke2-cluster#step-3:

• [x] Screenshot says "us-east-1" and I think we agreed that for now the workshop only supports that region, the default in Rancher UI is "us-west-2", no instructions if I should change the default location at this point? I assumed it's better to go with the value on the screenshot.

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/30-creating-a-cluster/32-creating-rke2-cluster#step-4:

• [x] Plan a break or something else in the workshop at this stage accordingly to bridge the wait time "The cluster creation process will take roughly 5 to 10 minutes."

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/40-importing-a-cluster/41-importing-the-cluster#step-5:

• [x] "Since there should not be any issues with importing the EKS Cluster, this will be the view of the EKS Cluster once it is imported into Rancher!" and if there is an issue? ;)

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/40-importing-a-cluster/41-importing-the-cluster#step-6:

• [x] Typo, "management clsuter" > "management cluster"

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/50-additional-integrations

• [ ] Couldn't follow along these steps because of the S3 access denied bug, it's not a mandatory workshop part so I'm fine with that as it is now, the steps make sense so this can be reviewed and improved at a later point and shouldn't block the release.

https://catalog.us-east-1.prod.workshops.aws/event/dashboard/en-US/workshop/50-additional-integrations#download-the-install-ack-sc-memdb.sh-script

• [ ] Second "Download the install-ack-sc-memdb.sh script" (Duplicate names) explains how to upload the file to cloud9, under the assumption I manually downloaded it somewhere, but it's not mentioned in the Workshop Environment section earlier? Also not on that page where to get this file.

[wombelix]

Hey @zackbradys I did it to give you more time, as stated above, just address the three major issues, let me know you did it and I will approve the Workshop. I trust you that you then going to work on the other content related issues as soon you have some more time to further improve the workshop :) No need for another review to help you get it done in time.

[zackbradys]

Hey @zackbradys I did it to give you more time, as stated above, just address the three major issues, let me know you did it and I will approve the Workshop. I trust you that you then going to work on the other content related issues as soon you have some more time to further improve the workshop :) No need for another review to help you get it done in time.

Thank you for replying! I re-read the issue after commenting and saw you mention that you didn’t need to do another review (thus why I deleted my comment). We will work towards getting these items addressed and send updates via this issue.

[marshkkm]

Hey @wombelix, the instructions were actually present in the markdowns...but as \

 blocks. So the workshop was ignoring them :). I've updated all of those instructions to a bulleted format.
            

[zackbradys]

Thank for the fixes @marshkkm and verifying the blockers are resolved. I added a bit of a formatting changes to it, but it looks good and checks off the 3 blocking tasks for approval.

I started another branch zackbradys-optional-updates-issue-40 to address the optional and non-blocking issues that you brought up. I agree with most of them so would like to get them implemented, but not even to hold off the workshop approval. Please let us know if we need to address anything else before approval!

[zackbradys]

Added pull request and merge for the optional suggestions for the Rancher components of the workshop. Please let me know any questions or additional suggestions!

[zackbradys]

For NeuVector, Step 1, nitpick, wouldn't it be easier to let people use the "Cluster Tools" link in the left menu to deploy NeuVector to ensure the Rancher integrated version is used?

I wanted to use the same method for deploying each application instead of having the user navigate to a different section. We could have both Longhorn and NeuVector be deployed from Cluster Tools, but it's splitting hairs at his point (IMO).

[zackbradys]

Please review the blocking updates made earlier tonight and let us know if they are acceptable for approval. We will continue to work on the optional/suggested updates. @wombelix

[wombelix]

Thanks a lot for addressing the findings that fast, as agreed, Workshop is now approved :)

All the best with the upcoming Event!

[zackbradys]

thanks for your help and working with us on our timeline. @wombelix!!