Query engine
flink
Question
I think flink dynamic iceberg sink support schema evolution. But it need DynamicRecord with Schema. What format of Schema should I provide？
e.g.
{
"schema-id": 1,
"fields": [
{"id": 1, "name": "user_id", "type": "long"},
{"id": 2, "name": "name", "type": "string"}
]
}
{
"schema-id": 2,
"fields": [
{"id": 1, "name": "user_id", "type": "long"},
{"id":2, "name": "email", "type": "string"},
{"id": 3, "name": "name", "type": "string"}
]
}
- should I provide a increment
schema-id?
- if I want to add a column, should I maintain the previous id sequence. or I can insert a column between previous column
like above(add a column "email", and not consider the id)
Query engine
flink
Question
I think flink dynamic iceberg sink support schema evolution. But it need
DynamicRecordwithSchema. What format ofSchemashould I provide？e.g.
schema-id?like above(add a column "email", and not consider the id)