# Case: tidbcluster not ready

Use when APPBox is Active and Cluster is Creating/Running but TiDB components
do not become Ready (infra/CR view — not deep TiKV performance).

## Steps

1. Confirm APPBox Active (hard Gate); else handoff `diag-appbox`.
2. Intent: DP `TidbCluster` (or v2) conditions/Ready (hard).
3. Intent: DP Pods ListByLabel for failing components (hard if TiDBCluster
   points at not-ready pods; else soft).
4. Soft: Pod events (`detail`+`events`).
5. If node shortage suspected and AWS/GCP/Azure CLI needed → user-assist
   (`skill-boundary-assist`) with filled cluster/nodegroup identifiers from
   evidence; keep placeholders for profile/account.
6. Code anchors when reachable:
   `cluster-controller/controllers/cluster/task/common/tidbcluster.go` (or
   `tidbcluster_v2.go`), `nodepool.go`, provider nodegroup tasks.

## After recovery

Health checklist items for TiDBCluster Ready + phase Running.