Maurício Fernandes de Castro (2) [Avatar] Offline
#1
My first contact with the toolkit you propose in the book was during Scala Days Berlin this year, where I quickly realized this is the path to follow. As my background is Java, JPA, Spring, etc., it took me some time to get a minimal experience with Scala, Play, and a still limited notion from Akka. Now that I started to actually rewrite code, I see that I need an intermediate step to keep things going before I can dive into Akka Persistence. I'd appreciate your advice to choose between two options: 1) replace JPA with Slick as a start, and wait until I am able to move to Akka Persistence, or 2) choose Akka Persistence directly even if book chapters I may eventually need would take some time to be released.

Thank you, and congratulations for the great work.
Sean A. Walsh (15) [Avatar] Offline
#2
Hi, question: Are you planning on going CQRS/Event Sourcing as well?
Maurício Fernandes de Castro (2) [Avatar] Offline
#3
Yes! Despite I do not have any previous CQRS experience to devise the full impact of such decision, it seems to me very appealing to separate strategies (and jvm's) to query and update the persisted state of a system.

In addition, the project I now plan to convert according to the "Reactive Manifesto" was open sourced in 2004 to incentivate best practices within our local ITC cluster. I expect to have other teams learning from my experience in a near future, including CQRS/Event Sourcing.

Thanks for the help,

Maurício Castro
Sean A. Walsh (15) [Avatar] Offline
#4
Maurício Fernandes de Castro wrote:Yes! Despite I do not have any previous CQRS experience to devise the full impact of such decision, it seems to me very appealing to separate strategies (and jvm's) to query and update the persisted state of a system.

In addition, the project I now plan to convert according to the "Reactive Manifesto" was open sourced in 2004 to incentivate best practices within our local ITC cluster. I expect to have other teams learning from my experience in a near future, including CQRS/Event Sourcing.

Thanks for the help,

Maurício Castro


Ok the reason I asked is that if you are planning on going the CQRS route then that migration to Slick doesn't really buy you anything in the long run since you will likely be using akka persistence to store events only, no more CRUD. Your migration path sounds like it would best be a full rewrite using akka persistence.

I would suggest taking a single area of your application or a single bounded context (domain driven design) and migrate that first. You application will eventually be a set of completely decoupled microservices.
nresni (8) [Avatar] Offline
#5
Hi,

I will begin a new project and I would like to be reactive ! smilie

You recommend to start directly with akka persistence, but in typesafe documentation, akka persistence is experimental, so my question is: Is it ready to go in production ?

Thanks for this book !

Nresni

Sean A. Walsh (15) [Avatar] Offline
#6
nresni wrote:Hi,

I will begin a new project and I would like to be reactive ! smilie

You recommend to start directly with akka persistence, but in typesafe documentation, akka persistence is experimental, so my question is: Is it ready to go in production ?

Thanks for this book !

Nresni



Nresni, the journal aspects of akka-persistence are as production ready as "experimental" gets in my opinion. What you should be careful of is how you model your CQRS read side as that part of akka-persistence is still evolving. I plan on rolling my own and simply publishing events and using listener actors on the read side as I've done with much success in the past, at least until the Akka team comes up with a better design.
nresni (8) [Avatar] Offline
#7
Thanks for your answer.