Services in Monoliths

[amithgeorge]

We have a monolithic Clojure (JVM) application. This application handles many different usecases. A usecase typically starts with a web request, which enqueues multiple background jobs. The jobs run and update the database.

In another forum post, I read a community member mention

> the customer, he can create a custom service for a method he is interested in and then we would capture purepaths for that method

Does this mean I can create custom services for certain usecases within my monolith? I find the idea interesting, because that would make it easy for me leverage builtin "Service-level availability" templates for the SLOs for that usecase.

[AntonPineiro · Accepted solution]

Hi,

You can create your own Custom services.

Best regards 

[Romanenkov_Al3x]

Custom service is great funcional.

This is how it looks:

After instrumentation Dynatrace will add service. You can find it in "called services".

in squares - additional information on the operation of methods instrumented using custom service

How exactly it takes - buildFeedMetadata method and buildFeedEntries

Great real example with monolith app:

we have successfully instrumented service, which generate insurance policy(PDF) and send it to the Customer (async logic).

Out of the box with Dynatrace we can see only http service - which send response 200(OK) in microseconds - for incoming generation request and other main work (for 30 sec+) and logic was hidden from us. 

We have added processing steps with Custom Service and have got:

if Service slowed Dynatrace exactly show which step of processing (Custrom Service) is the root cause of the problem.

Regards,

Alex Romanenkov