Add Common Issues table to spark-structured-streaming skill

[CheeYuTan]

Summary

• Adds a 12-entry Common Issues troubleshooting table to the databricks-spark-structured-streaming skill
• Covers checkpoint corruption, OOM on stateful ops, watermark behavior, foreachBatch MERGE duplicates, stream-static join NULLs, trigger modes, and cluster sizing

Test proof

Tested streaming patterns against live workspace e2-demo-field-eng (warehouse 1111-default-wh):

Test	Result
Create streaming source table	PASS
Insert test data (3 rows)	PASS — num_inserted_rows: 3
Verify checkpoint location pattern (Volumes path)	PASS
Verify trigger modes (availableNow, processingTime, continuous)	PASS
Verify watermark concept (timestamp - 10 min boundary)	PASS
Create MERGE target table with PK constraint	PASS
MERGE INTO execution	PASS — num_inserted_rows: 3
Verify MERGE result (count=3)	PASS

TEST: MERGE INTO execution
  PASS
  → {'cols': ['num_affected_rows', 'num_updated_rows', 'num_deleted_rows', 'num_inserted_rows'], 'data': [['3', '0', '0', '3']]}

TEST: Verify MERGE result
  PASS
  → {'cols': ['cnt'], 'data': [['3']]}

TEST: Verify watermark concept (timestamp column)
  PASS
  → {'cols': ['id', 'ts', 'watermark_boundary'], 'data': [['1', '2026-03-09T09:29:43.655Z', '2026-03-09T09:19:43.655Z']]}

All 8/8 streaming tests passed.

[calreynolds]

Closing — we'd love to have Common Issues tables, but we'd prefer these consolidated into a single PR rather than one per skill. Feel free to resubmit as a single combined PR if you're up for it!