processor or exporter？ which is better to send the resource information ingested from signals to kafka？

[zephyrpathsofglory]

Dear Everyone:

    I have received a requirement recently, that is send the Resource information, such as ResourceLogs.Resource, to a specific Kafka topic in a pipeline.

    IMO, an exporter is to send signals only, like kafkaexporter to send signals to Kafka, datadogexporter to send signals to datadog backend, if I want to send resource information to Kafka, does using exporter break the convention of OpenTelemetry? If I use processors, it seems that doing some write operations in a processor also breaks the convention of OpenTelemetry.

    Which is better? Or do you have any better alternative solutions? Please help, thanks