Add exist tools compare

010_replace_rabbitmq/readme.md

@@ -129,6 +129,21 @@ However, it is not trivial to implement all the communication needed from handsh
 The `kiwipy` in AiiDA servers as the role to provide the wrappered methods to talk between different components over RabbitMQ.
 Therefore, `kiwipy` is the interface for actioner and worker on bundle the operations to talk to RMQ, as the replacement, in `tatzelwurm` I should provide the python interface with methods ready to be used for sending and consuming certain type of MessagePack meesage.
 
+## Other solution or exist tools?
+
+It is discussed in https://github.com/aiidateam/AEP/pull/30#discussion_r813895745, better to look for tools that ready to use.
+What mentioned are [Apache Kafka](https://kafka.apache.org/), [faust-streaming](https://github.com/faust-streaming), [MQTT](https://mqtt.org/) and [Celery](https://docs.celeryq.dev/en/stable/getting-started/introduction.html).
+
+- Apach Kafka: for event streaming, require service to start. (servirce required)
+- MQTT: is a messaging protocol in low bandwidth environments. (no queue system + overkill for low bandwidth)
+- faust-streaming: require Kafka as event streaming backend. (Kafka services required, it is like kiwipy but for Kafka)
+- Celery: this is a task queue. (For event streaming it requires a message broker, RMQ, Redis, SQS .., For persistent, it requires a results store.)
+
+The `tatzelwurm` is tailored for AiiDA use cases specifically with message streaming/broke, lightwight queue system and persistent with KV store in disk in design.
+It is lightweight but will meet all the requirements for AiiDA.
+The tools above are for machines to machines scenario which required for huge message passing around millios of microservices and that is why those are overkill.
+We are not targeting to that but for AiiDA with required architecture as described above. 
+
 ## Design details
 
 In this section, I futher extend on the details of the new architecture.