Is your feature request related to a problem?
The current protocol=avro in TiCDC is highly efficient but only provides the After state of a row. While Data Engineering teams are satisfied, Backend Services require the Before state for auditing, stateful logic, and cache invalidation. Users currently have to choose between efficiency (Avro) and data richness (Debezium JSON/Canal JSON).
Describe the feature you'd like
When using avro format to kafka, add a query parameter like include-before-value, when it is true, an extension field _ticdc_before is added to the message. Like:
{
// --- Existing Root Level (The "After" state) ---
"id": 1001,
"name": "Howard",
"age": 20,
// --- New Optional Metadata Field ---
"_ticdc_before": {
"id": 1001,
"name": "Howard",
"age": 18
},
// --- Existing Metadata ---
"_ticdc_op": "u",
"_ticdc_commit_ts": 442318927364526081
}Describe alternatives you've considered
No response
Teachability, Documentation, Adoption, Migration Strategy
No response
Is your feature request related to a problem?
The current protocol=avro in TiCDC is highly efficient but only provides the After state of a row. While Data Engineering teams are satisfied, Backend Services require the Before state for auditing, stateful logic, and cache invalidation. Users currently have to choose between efficiency (Avro) and data richness (Debezium JSON/Canal JSON).
Describe the feature you'd like
When using avro format to kafka, add a query parameter like
include-before-value, when it istrue, an extension field_ticdc_beforeis added to the message. Like:{ // --- Existing Root Level (The "After" state) --- "id": 1001, "name": "Howard", "age": 20, // --- New Optional Metadata Field --- "_ticdc_before": { "id": 1001, "name": "Howard", "age": 18 }, // --- Existing Metadata --- "_ticdc_op": "u", "_ticdc_commit_ts": 442318927364526081 }Describe alternatives you've considered
No response
Teachability, Documentation, Adoption, Migration Strategy
No response