Questions on the use of event storming in application migration

56 views
Skip to first unread message

Jong Il Kim

unread,
May 22, 2021, 8:42:34 PM5/22/21
to EventStorming
My client has been in operation for 15 years. I want to re-develop the system as MSA.  Existing applications consist of a single database and are made of highly coupled codes.  I think it's almost impossible to analyze all of this code at once before developing it.  

So, I tried to identify the bounded context by using event storming and gradually analyze and deploy the existing codes.  This is because there is a difference in detail between the event-storming result and the actual event in the existing source code.  

However, the problem is that the customer wants the content from most of the existing sources to come out as a result of eventstorming.  

Does anyone have any experience with this?

Ryan Marsh

unread,
May 22, 2021, 10:50:50 PM5/22/21
to Jong Il Kim, EventStorming
What do you mean by "the customer wants the content from most of the existing sources to come out as a result of eventstorming". What is "the content" and "existing sources" in this context?

Do you mean the customer wants to model the current state of things in event storming.

Also, did you start with Big Picture event storming or did you dive right into Process/Software Modeling?

Kind regards, 
Ryan Marsh

Sent via Superhuman


--
You received this message because you are subscribed to the Google Groups "EventStorming" group.
To unsubscribe from this group and stop receiving emails from it, send an email to eventstorming+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/eventstorming/9bc3f690-6ad8-42d8-b979-fb61c3911a85n%40googlegroups.com.

Alberto Brandolini

unread,
May 26, 2021, 3:26:00 AM5/26/21
to EventStorming
Hi Nijin, there are quite a few strange things, so there will be more questions than answers...

Il giorno domenica 23 maggio 2021 alle 04:50:50 UTC+2 ry...@thestack.io ha scritto:
What do you mean by "the customer wants the content from most of the existing sources to come out as a result of eventstorming". What is "the content" and "existing sources" in this context?

Do you mean the customer wants to model the current state of things in event storming.

Also, did you start with Big Picture event storming or did you dive right into Process/Software Modeling?

Kind regards, 
Ryan Marsh

Sent via Superhuman


On Sat, May 22, 2021 at 7:42 PM, Jong Il Kim <nij...@gmail.com> wrote:
My client has been in operation for 15 years. I want to re-develop the system as MSA.  Existing applications consist of a single database and are made of highly coupled codes.  I think it's almost impossible to analyze all of this code at once before developing it.

Are you building a replacement system, or are you refactoring the exiting monolith into something new?

EventStorming is not for analyzing code, but it plays well in analyzing the friction between the business need and the existing code. 
I usually find very little value in understanding the monolith as a whole. While there's a lot of value in understanding the points where the legacy application is creating troubles to the existing business needs.

 

So, I tried to identify the bounded context by using event storming and gradually analyze and deploy the existing codes.  This is because there is a difference in detail between the event-storming result and the actual event in the existing source code.  

This is right, but the level of detail is not the only difference. From here it smells like EventStorming was used only to understand how the legacy worked. Less interesting than understanding how the legacy should work instead. Was the business actually involved in the exploration process?

However, the problem is that the customer wants the content from most of the existing sources to come out as a result of eventstorming.

Why? Is this a Technical need or a business need? Are they trying to blueprint the entire migration? This might not be the best strategy: EventStorming is a learning tool and it's better to keep the learning as close as possible to the implementation, but learning everything in a week now, to have a blueprint for migrating in 6 months is a waste. What are they expecting exactly?
 
 

Does anyone have any experience with this?

If you mean business stakeholders with weird expectations.... yes. But I guess you need to clarify the reasons behind the customer's request. Sometimes there is something they're not telling you, sometimes the need is right but the path is wrong, so you can propose another strategy to achieve the same result. Sometimes they completely got the tools wrong, and that will take time. Other times, when I understood what they really had in mind ...I realized that wasn't the right place for me to stay.

Alberto 
 

--
You received this message because you are subscribed to the Google Groups "EventStorming" group.
To unsubscribe from this group and stop receiving emails from it, send an email to eventstormin...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages