Message access requirements for message queues

Message queue function: serves as a buffer for components of different processing speeds
Message queue requirements
- Message preservation \
- Repeat message processing \
- Message reliability assurance

List-based message queue solution

Message preservation \
- List Natural Guarantee
- Write message lpush
- Read rPOP or BRPOP (block client wait notify, do not consume CPU resources)
Duplicate message processing
- Message queues should be able to provide globally unique IDS (snow Flake) \ for each message
- The consumer program records the ID number of the message that has been processed
- Ensure idempotency of consumption
Message reliability assurance
- BRPOPLPUSH inserts the message into another List backup List
Possible problem: Producers send messages quickly and consumers process them slowly, causing messages in the List to pile up and strain Redis memory
List does not support multiple consumers. Steam can be used instead

Streams is a Redis data type designed specifically for message queues (memory-based message queues) \
operation
- XADD: Insert message, ensure order, can automatically generate global unique ID\
- XREAD: used to read messages, which can read data by ID (set block value) \
- XREADGROUP: reads messages as a consumer group \
- The XPENDING command can be used to query for all messages \ that have been read but not confirmed by consumers within each consumer group
- The XACK command is used to confirm to the message queue that message processing has completed
- XGROUP Creates consumer groups

Memory message queues use Streams after Redis 5.0
Message queue requirements
- Message sequencing Sequential access to message data
- Duplicate message processing message data has a globally unique number \
- Message reliability Ensures that message data is deleted after consumption