ABAP and JAVA Architecture

1. Message Server:It handles communication between distributed Dispatchers in ABAP system.
2.  Dispatcher Queue: Various work process types are stored in this queue.
3. Dispatcher: It distributes requests to the work processes.
4. Gateway: It enables communication between SAP system and between SAP system and external systems.
5. ABAP-Work processes: – It separately executes dialog steps in R/3 applications.

   Types of work processes are given as below:-

   

   

6. Memory-pipes: It enables communication between ICM and ABAP work processes.
7. Message Server: It handles java dispatchers and server processes.It enables communication within java runtime environment.
8. Enqueue Server:It handles logical locks that are set by the executed Java application program in a server process.
9. Central Services: Java cluster requires a special instance of the central services for managing locks and transmitting messages and data. Java cluster is a set of processes that work together to build the reliable system. Instance is group of resources such as memory, work processes and so on.
10. Java Dispatcher: It receives the client requests and forwards to the server process.
11. SDM: Software Deployment Manager is used to install J2EE components.
12. Java Server Processes: It can processes a large number of requests simultaneously.
13. Threading: Multiple Processes executes separately in the background, this concept is called threading.
14. ICM: It enables communication between SAP system and HTTP, HTTPS, SMTP protocol. It means by entering system URL in the browser you can access SAP from browser also.

One more component is JCO. JCO is used to handle communication between java dispatcher and ABAP dispatcher when system is configured as ABAP+Java.

How the SAP Logon Process works?

Step 1) Once a user clicks on the SAP system from GUI, the user request is forwarded to Dispatcher.
Step 2) Request is stored in Request queues first. Dispatcher follows First in First out rule. It will find free work process and if available will be assigned.

Step 3) As per user request, particular work process is assigned to user. For example, when user login to the system then Dialog work process is assigned to the user. If user runs a report in background then background work process is assigned to the user.When some modifications are done at database level then update workprocess is assigned.So as per user’s action workprocess is assigned.

Step 4) Once user is assigned the dialog workprocess then user authorizations, user’s current setting are rolled in to work-process in shared memory to access user’s data.Once dialog step is executed then user’s data is rolled out from workprocess. Thus shared memory will be cleaned and other user’s data can be saved in shared memory area. Dialog step means the screen movements. In a transaction, when a users jumps from one screen to other the process is called a dialog step.

Step 5) First work process will find the data in the buffer. If it finds data in buffer then there is no need to retrieve data from database. Thus response time is improved and this process is called hit.If it does not find the data in buffer then it will find the data in database and this process is called miss. Hit ratio should be always higher than miss ratio. It improves the performance of system .

Step 6) Other requested data is queried from the database and once the process is complete,the result is sent back to GUI via dispatcher.

Step 7) At the end user’s data is removed from shared memory so the memory will be available to other users.This process is called roll-out.