0

In today's world, where code resides on ephemeral instances (e.g. serverless or containers), the connections to Amazon MQ would have to be constantly created and destroyed. Is there a performance overhead in such scenarios? Is there a connection pooling mechanism to get around such performance overhead?

Coming from a database back-ground, there is pgBouncer and ProxySQL that serve as proxy for PostgreSQL and MySQL. Is there something similar for Amazon MQ?

cogitoergosum
  • 2,309
  • 4
  • 38
  • 62

1 Answers1

0

I'm fairly uneducated on using Amazon MQ so excuse the ignorance, but I remember an interesting topic from last years re:Invent that popped into my head when I saw your question. API gateway support for websocket connections using Lambda behind your API gateway. The ws connection is maintained by API gateway taking the heavy lifting off of application servers or in the new world serverless.

A quick google search led me to this Integrating AMQ with Rest API gateway. The answer there points to using a Lambda function between API Gateway and AMQ.

This is also a good read. https://aws.amazon.com/blogs/compute/announcing-websocket-apis-in-amazon-api-gateway/

Michael Quale
  • 568
  • 3
  • 16
  • "The answer there points to using a Lambda function between API Gateway and AMQ." That means, the connection with AMQ is still from the Lambda. Therefore, same as my question, right? I get the point on API Gateway maintaining the WS connection; but not sure how that translates into maintaining the AMQ connection. – cogitoergosum Aug 14 '19 at 00:59