GCP PubSub Subscription Lag
GCP PubSub subscription accumulates backlog when subscriber throughput is insufficient.
Category archive
Published troubleshooting guides for messaging issues.
GCP PubSub subscription accumulates backlog when subscriber throughput is insufficient.
NATS cluster routes disconnect when network partition separates nodes.
ZeroMQ CURVE encryption fails when server public key does not match.
ZeroMQ drops messages when socket HWM is reached and Conflate is not set.
Artemis security denies access when role permissions are not defined.
Artemis paging store fills disk when address full policy is page.
Artemis cluster connections fail when bridge configuration is wrong.
SQS FIFO queue deduplication fails when content-based dedup is disabled.
SNS topic does not deliver to subscribers when endpoint is invalid.
SQS dead letter queue does not receive messages when redrive policy is wrong.
SQS consumers hit API rate limit when polling frequency is too high.
SQS SendMessage fails when queue policy denies principal access.
NATS clients cannot subscribe when authorization token is invalid.
NATS JetStream rejects messages when account storage limit is reached.
NATS Streaming cluster does not form when RAFT consensus cannot be reached.
Pulsar message dispatcher cannot deliver when consumers are disconnected.
Pulsar producers are blocked when topic backlog exceeds quota.
Pulsar topic bundles cannot be unloaded when broker is overloaded.
Pulsar brokers are unavailable when Zookeeper metadata is inconsistent.
Redis stream entries remain pending when consumers do not send XACK.
ActiveMQ KahaDB index is corrupted after unclean shutdown.
ActiveMQ loses database lock when lease period expires without renewal.
ActiveMQ marks consumers as slow when prefetch is not consumed in time.
ActiveMQ producers block when destination is full and policy blocks producers.
ActiveMQ failover does not work when shared storage is not accessible.
Kafka clients are denied access when ACL policies do not permit operations.
Kafka Cruise Control cannot rebalance when proposed movements exceed capacity.
Kafka MirrorMaker falls behind when source cluster throughput exceeds replication capacity.
Kafka ksqlDB queries return no results when stream or table is not properly defined.
Kafka Connect connector tasks fail when source or sink is unreachable.
Kafka Schema Registry rejects messages when schema compatibility check fails.
Kafka cannot delete old log segments when retention policy is blocked.
Kafka consumer group rebalance fails when coordinator is unavailable.
Kafka producer batches timeout when broker acknowledgment is not received.
Kafka consumer requests offset that is no longer available due to retention.
Kafka partition has no leader when all ISR replicas are unavailable.
Kafka partitions are under-replicated when broker replicas are offline.
Kafka broker loses Zookeeper session when heartbeat timeout is exceeded.
Kafka broker fails to start when Zookeeper is unreachable or log dirs are full.
Kafka consumer lag grows when consumption rate is slower than production rate.
RabbitMQ messages are lost after restart when queue is not durable.
RabbitMQ message TTL does not expire messages when queue has no consumers.
RabbitMQ dead letter exchange does not receive messages when DLX configuration is wrong.
RabbitMQ consumers are cancelled when queue is deleted or mirrored queue loses master.
RabbitMQ federation link is down when upstream server is unavailable.
RabbitMQ shovel plugin does not forward messages when destination is unreachable.
RabbitMQ mirrored queue synchronization is stuck when network is slow or queue is large.
RabbitMQ network partitions cause split-brain when cluster nodes lose connectivity.
RabbitMQ blocks publishers when memory usage exceeds high watermark threshold.
RabbitMQ messages accumulate when consumers are disconnected or prefetch limits are reached.