Question
NiFi Publish Kafka hashes partition info automatically
Hello everyone,
I want to use the Publish Kafka processors and I need to specify the partition for each flow file. But when I choose Expression Language Partitioner class and add my partition info on Partition field, it does not publish my records to appropriate partitions. I realize that, instead, it gets my partition string and hashes. Then it publishes the records to partition with hashed value.
I dont understand why it works like this. I want to choose a spesific partition for each flow file. How can I achieve this?
Here is the code from nifi github repo. As we can see it hashes the partition value:
private Integer getPartition(final ProcessContext context, final FlowFile flowFile) {
final String partitionClass = context.getProperty(PARTITION_CLASS).getValue();
if (EXPRESSION_LANGUAGE_PARTITIONING.getValue().equals(partitionClass)) {
final String partition = context.getProperty(PARTITION).evaluateAttributeExpressions(flowFile).getValue();
final int hash = Objects.hashCode(partition);
return hash;
}
return null;
}
