Remove usages of findAll for types supported by server-side pagination

[richard-cox]

Blocked on #10786

Summary

Contributes to #9964

Occurred changes and/or fixed issues

• Where findAll is used try to instead
  • A) fetch a smaller set of resources via existing API methods (like filtering on namespace)
  • or
  • B) Performance Setting on - Fetch via new API filtering. Setting off - continue as is
• This code wise is based on top of Wire in pagination to LabeledSelect and use for pods, secrets and configmaps #10786 which does a similar thing but for findAlls rather than LabelSelects

In Addition

• With the steve cache enabled the Workflow Workloads list should not be navigatable. This will be replaced at some point with an overview page
• Fixed Rancher Backups --> Backups, Resource --> Create/edit page bottom bar (was not stuck to bottom)

Technical notes summary

• There's two temp bools in shell/utils/pagination-utils.ts, TEMP_VAI_CACHE_MERGED and TEMP_PERF_ENABLED. These will be removed pre 2.9.0 when the steve cache is merged and we can just run on the FF and perf setting to enable/disable the features
• For this PR would recommend
  • Provision Rancher with the vai image (available on request)
  • Enable the steve cache Feature Flag
  • Enable pagination via the Performance setting

Areas or cases that should be tested

• PR wise with the new vai cache image and everything enabled. In addition maybe a bit with it disabled. QA should be looking at more thorough tests as per below

Areas which could experience regressions

With both the performance pagination setting and feature flag off

• Harvester Machine Config gets correct config maps. Cluster Management --> Cluster list --> Create --> RKE2 Harvester --> User Data Template and Network Data Template contains correct config maps
• New ResourceLabeledSelectcontinues to work. Nav tolocalcluster -->Cluster Tools-->Rancher Backup Install--> Next -->Use an SE-compatible object store-->Credential Secret should show values
• SelectOrCreateAuthSecret continues to work. Nav to local cluster --> Apps --> Repositories --> Rancher row Edit Config --> http(s) URL to an index generaed by Helm --> Authentication user can create either type and also select an existing one
• Cloud Credential API Key field correctly populates. (cloud creds should exist, type shouldn't matter ) Cluster management --> Cloud Credential --> API Key column should be populated
• Ingress target should load correctly. Nav to local cluster --> Service Discovery --> Ingresses --> The Target columns should contain /
• Ingress workflow detail page shows Certificate link. Nav to local cluster --> Service Discovery --> Ingress --> click on rancher --> Rules tab --> confirm Certificate column is populated
• Ingress create/edit Certifcate secret selection works correct. Nav to local cluster --> Service Discovery --> Ingresses --> Create --> Certificates tab --> Certificate - Secret Name contain at least one option Default Ingress Controller Certificate and changing the namespace to kube-system should then reveal another option (depends on kube flavour, on k3s its k2s-serving).
• Node Pods list is populated. Nav to local cluster --> Cluster --> Nodes --> click on any node --> Pods tab --> Confirm the pods list is populated
• Backup Encryption key can be selected. Nav to local cluster --> (install Rancher Backups) --> Rancher Backups --> Backups --> Create --> Encryption --> Select Encrypt backsps using... --> Drop down below should be populated (conditions for secrets in tooltip must be met)
• Restore Credential secret can be selected. Nav to local cluster --> (install Rancher Backups) --> Rancher Backups --> Restore --> Create --> Select An S3-compatible bject store --> Drop down below should be populated
• ServiceAccounts secret can be selected. Nav to local cluster --> (ensure a secret of type kubernetes.io/dockerconfigjson has been created in a specific namespace) --> More Resources --> ServiceAccounts --> Create --> select specific namespace --> Service Account tab --> Pull Secrets should contain secret. Changing the namespace should result in the correct secrets relative to that namespace
• Project Montioring CRUD is not broken. Nav to local cluster --> (Install monitoring and prometheus federator apps) --> Monitoring --> Project Monitoring --> can succesfully create and view one

The above should be repeated with the performance pagination setting and feature flag on. In Addition

• With the steve cache enabled the Workflow list should not be navigatable. This will be replaced at some point with an overview page

Checklist

• The PR is linked to an issue and the linked issue has a Milestone, or no issue is needed
• The PR has a Milestone
• The PR template has been filled out
• The PR has been self reviewed
• The PR has a reviewer assigned
• The PR has automated tests or clear instructions for manual tests and the linked issue has appropriate QA labels, or tests are not needed
• The PR has reviewed with UX and tested in light and dark mode, or there are no UX changes

[richard-cox]

@codyrancher @nwmac I'm having fun getting this one passed the gates (some might be geniune, but a lot of other prs are currently failing). The code should hopefully remain unchanged, would it be possible to review now to save time?

[codyrancher]

I didn't verify Harvester or S3 due to access.

From what I could verify it looks like we're still populating resources just fine.

Do you have a script or something for generating the resources in the future? It's fairly time consuming testing something like this.

[richard-cox]

@codyrancher Yeah there's a few requirements on certain types of secrets. I'm aiming to have some e2e tests where we can upload resources via YAML to populate pages.