Change Log for the Integration Content: KaTe Kafka Adapter for Integration Suite - Version 1.0.16: August 2026 New feature: Sender channels now also emit the iFlow message properties kafkaKeySchemaId and kafkaValueSchemaId for deserializers that use Schema Registry. This allows routing by different schema IDs ("versions") in an iFlow. Sender channels also support filtering by these fields using EQ with existing _OR_ and _AND_ composition. New feature: Sender batches now allow control over which metadata is added alongside key and value: record position metadata, Kafka headers, and schema IDs, with backward-compatible defaults. New feature: New error handler option: In addition to the existing error-topic/DLQ option, sender channels can now forward failed records to a dedicated internal error-handler iFlow, where customers can process the original Kafka record using their own handling logic, for example by storing it in a JMS queue. Fix: Batch sender error handling now places the failed aggregate MPL ID in the cpiMessageId header of every record forwarded from that batch, allowing error-topic records to link directly back to the FORWARDED monitoring entry. Fix: Improved SAP PO migration and backward compatibility for nullable XML-to-Avro union values. - Version 1.0.15: August 2026 New feature: Added Confluent Client-Side Payload Encryption (CSPE) for Avro, JSON Schema, and Protobuf values in both receiver and sender channels. Encryption and decryption use native `ENCRYPT_PAYLOAD` Schema Registry rules and fail closed on errors. New feature: Added atomic receiver batching for simple XML, simple JSON arrays, and detailed XML envelopes. A complete batch is published in one Kafka transaction and becomes visible atomically to `read_committed` consumers. New feature: Sender channels can select `read_uncommitted` (the compatibility default) or `read_committed` transaction isolation. New feature: SAP Cloud Connector support, SCC connections without compromising TLS hostname verification or extra listener exposure on internal Kafka brokers added. Updated: Kafka Clients were upgraded to 4.x a - Version 1.0.14: July 2026 New feature: Tombstone records can now also be set not only on via channel configuration level but also with the `isTombstone` header. Accepted values are `true`/`false`, `tombstone`/`data`, `delete`/`record`, `yes`/`no`, and `1`/`0`. Fixed: batch sender channels had recovery problems on rebalance, this is now fixed with this release. - Version 1.0.13: June 2026 Fixed Kafka-192: XML2AVRO now converts enum values inside Avro arrays to Avro enum symbols instead of plain strings. Fixed Kafka-193: XML2AVRO now accepts omitted XML elements for required Avro arrays and writes an empty Avro array. Nullable arrays still remain `null` when the XML element is omitted. - Version 1.0.12: June 2026 - Fixed wrong conversion of decimal logical Avro types - Version 1.0.11: June 2026 New feature: Support for Confluent Client-Side Field-Level Encryption (CSFLE) on payloads in both sender and receiver channels. When the schema registered in the Confluent Schema Registry carries a `RuleSet` with an `ENCRYPT` rule (e.g. tagging fields as `PII`), the adapter now transparently encrypts tagged fields on receiver (produce) channels and decrypts them on sender (consume) channels using the DEK returned by the Schema Registry. - Version 1.0.10: May 2026 Fixed the spelling of the Kafka monitoring field `isThombstone` to `isTombstone`. The exchange property / header set on tombstone records (empty Kafka data) is now emitted under the correct, documented name `isTombstone`. iflows reading the previous misspelled property must be adjusted accordingly. Improved readability of license check error messages. New feature: Configurable startup connectivity checks. Four new options (`Ping on startup`, `Ping bootstrap servers`, `Ping topic`, `Ping schema registry`) on the *Extended* tab of sender and receiver channels let you en/disable the checks performed when a channel starts. The master switch `Ping on startup` (default on) gates all three sub-checks; each individual check (default on) only runs when the master is enabled. The schema registry check additionally requires that schema registry usage is configured. Version 1.0.9: April 2026 Fixed: the `retryCount` value exposed to the iflow now honors the channel's *Kafka headers/properties* setting (`Only headers` / `Only properties` / `Both`). Previously it was always set as a Camel header, which made it invisible to iflows configured for `Only properties`. Documentation: new how-to *Designing error handling for sender channels* with five recipes covering default retries, reacting to retries via `retryCount`, automated *ignore* / *send to error topic* handling, and the combined production setup. Version 1.0.8: April 2026 Changed default error behavior for non-custom-handled sender errors so the adapter creates error exchange entries or message processing logs instead of surfacing them via the iflow deployment status. Version 1.0.7: March 2026 Added non-validating and validating JSON Schema serializer/deserializer options alongside the existing plain JSON serializer/deserializer Schema registry lookup for Confluent serializers (Avro, JSON Schema, Protobuf) now uses schema ID instead of version when pinning a specific schema. Custom Avro header serializer continues to use version-based lookup. Version 1.0.6: Februar 2026 Fixed wrongly spelled custom status for ignored Fixed problems with kafka keys rendered as binary string Version 1.0.5: October 2025 Fixed issues for publishing tombstone records. Also Avro serialized keys are now possible for tombstones. Version 1.0.4: September 2025 Added new options for fine grained serialization configuration Version 1.0.3: September 2025 Added support for confluent json schema serializer & deserializer with schema registry Version 1.0.2: August 2025 Upgrade to Kafka 3.9.1 clients The adapter now supports Oauth bearer authentication for broker and service registry connections Version 1.0.1: July 2025 New feature: Aside the Confluent Avro serializer/deserializer, the adapter now supports now also Header based serialization/deserialization for Avro (whereas schema name/version are stored in headers instead of a byte in the data section. Minor fixes for monitoring/error handling Version 1.0.0: April 2025 Initial Release of KaTe Kafka Adaptor