> GCP has done well integrating their different services together. GCP provides a smaller set of core primitives that are global and work well for lots of use cases. Pub/Sub is probably the best example I have for this. GCP has Pub/Sub. Pub/Sub seems flexible enough to replace most (all?) of AWS’ various queues.
I feel like if you're going to spend this much time writing something in this much detail, the least you can do is try to understand some of the things you're criticizing and not say things so wildly incorrect.
> In AWS you have SQS, SNS, Amazon MQ, Kinesis Data Streams, Kinesis Data Firehose, DynamoDB Streams, and maybe another queueing service by the time you read this post.
The most accurate criticism would be that Pub/Sub combines AWS's SQS and SNS. AWS had SQS first for 1-to-1 async queuing integrations. Then they added SNS for fan-out. GCP just started with fan-out from the start and said "well if you don't want to fan out, and just have one subscription, it's basically fine.
That's actually also true with AWS too - nobody is stopping you from using SNS this way. SQS was the first AWS service (S3 went "GA" first, though), and AWS doesn't deprecate things that are popular with customers.
Amazon MQ is Amazon's managed Apache MQ. You can criticize Amazon for hosting managed open source software, but there is a reason why it exists for customers that want to use an industry standard queueing service but not worry about operating it.
GCP's competitor to Kinesis isn't Pub/Sub, it's Dataflow -> https://cloud.google.com/dataflow Both are described as for the same use case - streaming, analytics, batch processing.
Kinesis Data Firehose may be another valid criticism. It can be thought of a "Simplified Kinesis that can push data instead of pulling it". It is clearly coupled tightly with Kinesis Data Streams itself, so in another company that cared less about not breaking backwards compatibility maybe they would've bolted those features on top of Kinesis.
I feel like if you're going to spend this much time writing something in this much detail, the least you can do is try to understand some of the things you're criticizing and not say things so wildly incorrect.
> In AWS you have SQS, SNS, Amazon MQ, Kinesis Data Streams, Kinesis Data Firehose, DynamoDB Streams, and maybe another queueing service by the time you read this post.
The most accurate criticism would be that Pub/Sub combines AWS's SQS and SNS. AWS had SQS first for 1-to-1 async queuing integrations. Then they added SNS for fan-out. GCP just started with fan-out from the start and said "well if you don't want to fan out, and just have one subscription, it's basically fine.
That's actually also true with AWS too - nobody is stopping you from using SNS this way. SQS was the first AWS service (S3 went "GA" first, though), and AWS doesn't deprecate things that are popular with customers.
Amazon MQ is Amazon's managed Apache MQ. You can criticize Amazon for hosting managed open source software, but there is a reason why it exists for customers that want to use an industry standard queueing service but not worry about operating it.
GCP's competitor to Kinesis isn't Pub/Sub, it's Dataflow -> https://cloud.google.com/dataflow Both are described as for the same use case - streaming, analytics, batch processing.
Kinesis Data Firehose may be another valid criticism. It can be thought of a "Simplified Kinesis that can push data instead of pulling it". It is clearly coupled tightly with Kinesis Data Streams itself, so in another company that cared less about not breaking backwards compatibility maybe they would've bolted those features on top of Kinesis.
DynamoDB Streams is also not a separate service. It's a feature of DynamoDB - streams data capture streams from a database. Exactly like... https://cloud.google.com/bigtable/docs/change-streams-overvi...
> 2019 Update: Amazon has now released another streaming service: Amazon Managed Streaming Kafka.
GCP also has a Managed Kafka service only theirs is a collaboration with Confluent: https://cloud.google.com/learn/what-is-apache-kafka
"Confluent Cloud on Google Cloud provides fully managed Apache Kafka as a service so you can focus on building apps and not managing clusters."