docs: add distributed XGBoost on Kubernetes tutorial (#12080)

Add comprehensive documentation for distributed XGBoost training on
Kubernetes via Kubeflow Trainer. The tutorial covers:

- Overview of the distributed architecture and Collective protocol
- Environment variables (DMLC_*) injected by the XGBoost runtime plugin
- Worker count calculation for CPU and GPU training
- Prerequisites and installation verification
- ClusterTrainingRuntime configuration
- Example training code using Python SDK and kubectl YAML
- Best practices: QuantileDMatrix, early stopping, checkpointing,
  logging, data partitioning, and rank-specific logic
- Common issues and edge cases

Signed-off-by: krishna-kg732 <krishnagupta.kg2k6@gmail.com>

Author: Krishna-kg732