How do you configure Spring Kafka consumer groups for multi-tenancy?

Kafka delivers each message to one process in each consumer group, so separate consumer groups give separate consumers independent progress. A common naming convention for tenant isolation is group.id = tenant-{tenantId}-{service-name}, set per listener with the groupId attribute. For tenants that have their own topics, a @KafkaListener can subscribe by regular expression with topicPattern, for example @KafkaListener(topicPattern = 'tenant-.*-orders'), which maps to the consumer's subscribe(Pattern); matching topics are checked periodically against the topics existing at the time of check (governed by metadata.max.age.ms), so a new tenant topic is picked up on a later metadata refresh. missingTopicsFatal does not apply to pattern subscriptions. Another approach is one shared topic where the listener routes each record to tenant-specific processing, for example by key or header. For complete isolation with a separate cluster per tenant, define multiple ConsumerFactory/ProducerFactory beans with tenant-specific configuration and multiple KafkaListenerContainerFactory beans, then select one with the containerFactory attribute.

Expected key concepts: group.id, tenant isolation, topicPattern, metadata.max.age.ms, KafkaListenerContainerFactory, containerFactory, ProducerFactory, ConsumerFactory