General Usability Issues

[richard-cox]

This is a placeholder/container issue to track bugs, UX, etc issues discovered during hackweek

Low value bugs (fixes should be straight forward)

• (BUG) Rancher Backups / Backups / Create - Schedule placeholder suggests @midnight yet when entered we say this is an Invalid cron schedule
• (BUG) Persistent Volume Detail Page (default Config tab)- Border box showing the associated PVC grows crazy high when the 'Customise` tab is selected
  • 
  • 
• (BUG/UX) Detail Masthead - Age field is not live
  • Screenshot taken many minutes after creation. Given that our table age is live, this seems a bit misleading
  • 
• (BUG) Setup Page has no vertical scroll bar
  • Note - Would Sean's changes to a responsive log in page fix this?
  • 

High Value UX

• (UX) K9S Comparison - There's no way to see an overview of which pods are consuming high CPU/Memory. This is useful when your tiny node can't hack it, made worse when it can't handle monitoring app. The info is shown in k9s, but didn't see it in kubectl get pod -n cattle-system -o wide
• (UX) Resource Finalizers - It's not obvious if a resource has a finalizer. We should make this more prominent (it doesn't have to be in-your-face obvious). As a secondary issue we should then think about how to help the user remove this, as a last resort, to remove a resource. This would help in situations where a cluster was now getting deleted.
• (UX) Cluster Tools List - When one is installed we show associated resources. If you click on the card it takes you through to app details page with the Resource list showing by default. If a NS filter is set (like the default) the resource list is empty (which jars when we've just said there are 'x' resources. Generally should detail page resource lists be fitlered by the NS values?
• (UX) Mistakes when installing helm apps are costly
  • If a mistake leads to a condition that will never succeed the user has to wait 10 minutes for the helm operation to fail before deleting/updating the helm app with a fix
  • There are helm operation pods in progress... but it's hard to align when multiple operations are running
  • (i tried to install a chart requiring a PVC, then made a mistake when installing longhorn)

Medium Value UX

• (UX) Cluster Tools List - Clicking on Install takes user straight to install wizard. I wanted to learn more before installed (via chart detail page)
• (UX) Create Custom Cluster / Cluster Details Page / Registration Tab - Confusing text shown
  • Unclear what this refers to, specifically when the command isn't shown (windows nodes can't initially be applied)
  • Suggestion - change both references to Run the command below....
  • 
• (UX) Install/Edit/Upgrade Chart Step 2 (yaml)
  • This page really needs an 'overrides' view. Currently you have to find values in default template
  • If an error occurs user has to restart wizard (longhorn sets itself as default... without unsetting existing default... causes helm error)
• (BUG) Explorer side nav highlight weirdness
  • Starred group is highlighted when workload is selected
  • Item is double highlighted (not possible to work around)?

General (known, issues somewhere)

• (UX) Creating namespaces in the same context as create project would be very handy
• (UX) Explorer side nav - Collapsing groups are a pain when switching back and forth between two resources

[catherineluse]

Can this be separated into different issues for triage purposes?

I don't think any of these items already have issues except for the last one, which is tracked here #6658

[richard-cox]

They can be split out when we commit to them

[catherineluse]

I'm saying they'd be more likely to get done if we didn't have to commit to all of them at once.