☂️ Job Page Redesign

[philrenaud]

Figma
PRD

Service Jobs

• Service Job: "Running" state status panel #16133
  • Service Job Panel: Make current/historical a query parameter #16340
  • Show a "Latest deployment summary" cell for service jobs in steady state #17240
• Service Job: Deployment status panel #16350
  • Service Job Deployment Panel: History and Update Params #16512
  • [ui, deployments] "Failed" status in a deployment panel: only show if "permanent" state (wont restart/resched) #16961
  • [ui, deployments] Remove the old Active Deployment panel #17191

System Jobs

• Set up steady-state panel in system jobs #17036
• [ui, deployments] handle node read permissions for system job panel #17073
• Look into whether a job.numDesiredAllocs property is feasible #17099

Batch Jobs

• Running Allocations denominator for batch jobs should combine all non-complete allocs #17139

Other Job Types

• [ui, deployments] Set up steady state panel in batch jobs #17114
• [ui] Set up steady state panel in sysbatch jobs #17242
• [ui, deployments] Set up steady state panel in Periodic/Parameterized jobs #17216

Status panel, generally:

• [ui] Sort versions ascending numerically, with "pending" at the end #17149
• [ui, deployments] Failed or Lost cell #16941
• [ui, deployments] Linking allocblocks and legends to allocation / allocations index routes #16821
• [ui, deployments] Update Params: convert "500000" to "50s" etc. #16881
• [ui] Job status/deployment panel: split allocation status into clientStatus and check health #16724
• [ui, deployments] link blocks / legends to their allocation / allocations index with filter #16802
• [ui, deployments] Add a "working" indicator (shimmer, or ...) to Unplaced and Pending alloc indicators #16797
• Consider refactoring allocation-status-block (and represented-alloc?) into classes per [ui] Splitting deployment allocs by status, health, and canary status #16766 (comment)
• [ui, deployments] Have a "Show latest deployment task events" toggle for steady state #17159
• [ui, deployments] Bug: "Failed" allocs of differing deploymentStatus.Healthy should not group together #17011
• [ui, deployments] Bug fix: alloc ordering to show should follow running/pending first, then count backwards by version #17058
• [ui, deployments] Tooltip showing group name and alloc_id on alloc box hover #17513
• [ui, deployments] fix: deployment history (left line on top of some boxes, or sometimes the line disappears) #17523
• [ui] "Promote Canary" should be inline with allocs, and only show up when all canaries are healthy #17525
• [ui, deployments] keyboard shortcuts for “Fail Deployment”, “Promote Canary” #17567
• add version # next to the new allocations when deploying #17627
• [ui, deployments] Job status panel legend: only show the # of health unknown for running allocs #17565
• [ui, deployments] Steady state: top level status shows if job is degraded #17597
• [ui, deployments] Don't link to legend items with 0 allocs #17643
• Alignment and spacing changes to the job status UI #17683

Misc:

• Mirage Deployment simulator
• Remove obsoleted components (job-client-status-summary)
• [ui, deployments] Why aren't child jobs watching allocations? #17193
• Add a "Will not Reschedule" filter to the job/allocations page #17221

Jobs Index page

• [ui] Jobs Index page: show "allocations running" in addition to the existing chart type #17283

[mikenomitch]

Note: An allocation that is completed and the rescheduled should be marked as rescheduled