ISM 웹 콘솔 - 인터페이스 개발

The flow of an interface development¶

• First make pieces of an interface. The pieces are field, field group, server, application, service, mapping, and etc.
• Then assemble pieces into a bigger component like data structure, system, and service model.
• Finally combine components into an interface.

This chapter explains each piece and component according to the flow.

Refer to tutorials(Realtime Message-DB tutorial) for handson guide.

Data¶

Input/output data structures of interfaces are managed here. Data structure has 3 types.

• ISM type
• consists of 1 or more field groups.
• requires fields and field groups.
• XML type
• normal xml message
• generated through sample xml data
• ISO type
• ISO 8583 message

Field

Field is the smallest unit of a ISM data structure. Field is like a column of a table, an attribute of a C struct.

Field has some properties to be used for transformation, validation, and formatting.

Field group

Field group is an ordered set of one or more fields. Field group is basic unit of ISM data structure.

Common

Common tab defines general properties of a field group.

Mapping

Mapping is to assign fields into a field group, to specify usage for each field in a field group.

Properties of mapping table are followings.

Key

Key is used for DB operation. One field group can have 0 or more key fields like a table. Key fields are used to generate condition parameters for

1. select ... where param1 = key_field_N and param2 = key_field_M and ...
2. update ... where key1 = key_field_N and key2 = key_field_M and ...
3. delete ... where key1 = key_field_N and key2 = key_field_M and ...

SQL Function

SQL Function is used during transformation. ISM generates query for target operation during transformation in batch/deferred interface. When ISM generates queries, it checks whether a field is used as a sql function. If the field is defined as a sql function, the result of transformation of that field is added to the generated query as a constant.

For example,

• Col3 is normal field not used as sql function.
• Mapping to that field assigns SYSDATE as default value.
• Transformation result of that field is SYSDATE.

ISM will generate this query.

insert into target_table(col1, col2, col3) values(?, ?, ?);

if query is literal type, generated query will look like this.

insert into target_table(col1, col2, col3) values('col1_value', 'col2_value', 'SYSDATE');

• Col3 is used as sql function.
• Mapping to that field assigns SYSDATE as default value.
• Transformation result of that field is SYSDATE.

ISM will generate this query.

insert into target_table(col1, col2, col3) values(?, ?, SYSDATE);

if query is literal type, generated query will look like this.

insert into target_table(col1, col2, col3) values('col1_value', 'col2_value', SYSDATE);

Null

If input value of Nullable field is null, ISM will do

• For realtime flat message
• fill filler character to the field
• For database operation
• set null parameter
• For file operation with fixed length
• fill filler character to the field
• For file operation with delimiter
• set empty value

Add a field

Go to mapping tab.

Click add field button(), a new row is appended at the end of the field list.

Click empty cell and popup field list dialog.

Select a field, and selected field is set to the new row.

To insert a field into a specific index, set index value and click add field button().

To delete a field, select a field and click delete button()

Excel upload

For the layout of excel file, refer to this file field group sample. This file is included under ISM_INSTALL_DIR/data.

Data structure

ISM

ISM data structure defines

• layout of data
• ISM data structure consists of at least 1 master field group.
• 1 master field group can have 0 or more detail field groups.
• Refer to File-DB Batch tutorial for how to generate ISM data structure.
• record delimiter
• Used for file based data
• repeated group and repeat count
• Only detail group can be repeated.
• Repeat count is not fixed, repeat count must be assigned somewhere in the data structure before that detail group.
• Refer to Message-DB Realtime tutorial for how to define repeat group and repeat count

XML

XML data is generated from sample xml.

ISM does not provide data structure generation from xml schema or DTD.

Refer to Message-DB Realtime tutorial for how to generate XML data structure.

ISO

System¶

System is a combination of application and server. One system can have one or more combinations of them. Usual production systems run same services on multiple servers.

Application

Application defines type of software or package.

Common

Common tab defines application type and account information(userid/password).

File application

ISM connects FTP or SSH daemon to get or put file(s).

• FTP
• sFTP
• FTPs
• SCP
• Tandem FTP
• SFTP(Windows)
• This type is to simulate FTP ascii type transfer in sFTP.
• Custom
• When

DB application

ISM connects database servers via JDBC protocol. When DB application is used as a target and XA transaction is required, ISM uses platform provided adapter or datasource to communicate.

• Oracle
• MSSQL
• Mysql
• DB2
• DB2AS400
• Informix
• Sybase
• Teradata

Web application

Defines Web server. ISM just calls http(s) url regardless the type of http application - php, asp, jsp, etc.

EJB application

ISM communicates with J2EE servers via two protocols. One is JMS and the other is EJB. ISM uses Web application type to communicate other than JMS and EJB.

JMS type

• ActiveMQ
• SonicMQ
• MQ

EJB type

• Websphere
• WebLogic
• JBoss-as-5
• JBoss-as-7

Message application

Message application is used to communicate with applications in which flat messages are used. Flat message usually have fixed length data.

These classification don't have different implementation. They just indicate what applications are running.

• Socket
• TP Monitor

Etc

Etc type has no specific properties. It indicates that application is not in the application list.

Server

Server defines physical host address or hostname.

System

System combines one application with one or more servers. ISM invokes service running on one of the servers against load balancing rule or constructs connection information like FTP or database.

Common

Common tab defines general policies for that system. Load balancing and status management are defined.

Auto recovery

Both Health check and auto recovery are on, then ISM tries to send dummy message for health check to that down server. The dummy message for health check is provided by user program. In ISM realtime ISM requires realtime parser. Realtime parser is used to extract information from input message and set value to outbound message. Another things parser should provide are ACK message and health check messages.

ACK message does not mean network level protocol but business level protocol. ACK is required, ISM ask parser to generate ACK message based on input message. ISM returns that ACK message to sender/source system.

Refer to Realtime parser for how to generate those messages.

Server Mapping

Here ISM combines application and server and assign special properties to that combination related with load balancing policy.

Click Add button() to add a new combination.

Click Application id or Application name cell to choose an application.

Click Server id or Server name cell to choose a server.

All the other properties of this combination are about load balancing policy.

Is master

This property is used when load balancing type is either

• failover
• dedicated

failover

• All the requests goes to the master server except only when master server is down.
• When master server is down, ISM tries to determine which is the backup server of this server.
• If all the servers are set as master, first server in the list is the master server.

dedicated

• All the requests goes to the master server even if when master server is down.

weight for load balancing

This property is used when load balancing type is

• Round robin

Default weight among the servers are same. Other value than 0 is assigned to a server, ISM sums up all the values and calculates weight per server.

Server status

Refer to Server status for values for available server status.

Default status value is No run.

• No run
• means status is not checked.

To enable server status check, set status value to Run.

Healthcheck interval

This value is used only when that server is down.

Interval means with which interval seconds ISM will try to check the status of that server.

Backup application index

This property indicates next server to invoke when this server is down. If the index is not assigned, i.e. -1, then ISM look for next master server from the list.

Proxy system

System id of the proxy system is meaningful. Proxy system means what common process will be used to communicate to real target system for what type of interface. Refer to Types of proxy system for usage of proxy system

Common system

Common system stands for a process which is in charge of communication to multiple systems. Common system id is the only meaningful property.

Interface¶

Service

Service is used to define what to do for source/target respectively.

For example, in a DB-DB batch interface, a service for source interface gets data from source system. Service definition for source interface contains query for getting data from source table. Service definition for target interface contains operation - CUD - and table name.

Common

Common tab defines general policies for that system. Load balancing and status management are defined.

DB Service

DB Service tab defines table related operations.

CRUD

• select
• insert
• update
• delete
• Insert if update failed
• insert if update count is 0
• When the record does not exist, insert that record.
• update if insert failed.
• update if insert raises duplication error.
• When the record already exists, update that record.
• insert(skip duplicate)
• Even if duplication error occurs, just ignore.
• update(skip count 0)
• Even if no record is updated, just ignore.
• delete(skip count 0)
• Even if no record is deleted, just ignore.
• dynamic
  • ---
• dynamic(skip error)
• ---
• user sql
• ---
• user sql(skip error)
• ---

Filter type

• Value
• ---
• Pattern
• ---
• Handler
• ---

Stored procedure type

ISM assumes query entered in the Query area as stored procedure call statement when Stored procedure property is checked. Query for stored procedure looks like this.

{call GetInterfaceWithOutput(#parameter_name#, ?, ?, ?, ?, ?)}

• #iparameter_name#
• Enclose parameter with #.
• parameter_name must match a field name of input data structure.
• ?
• ? means input or output parameter.

ISM replace #parameter_name# with ? if query type is prepared. It means that prepared type query is used, ? can be used for input parameters.

Data structure of stored procedure

• stored procedure has no input/output parameters and returns result as a result set.

Contents of stored procedure(mysql)

DELIMITER //
DROP PROCEDURE IF EXISTS GetAllInterfaces;
CREATE PROCEDURE GetAllInterfaces()
    BEGIN
    SELECT 'GetAll', INTEGRATIONSERVICEID, ISTWOPHASECOMMIT, INTEGRATIONSERVICENAME, INTEGRATIONSERVICEDESC FROM integrationservice;
    END //
DELIMITER ;

ISM query

{call GetAllInterfaces()}

Input data structure is not required.

Output data structure

IN or OUT type value is not checked. Output data structure must contain one master field group with RESULT_COUNT field only.

• Result set data are mapped as detail field group.
• stored procedure has only input parameters and returns result as a result set.

Contents of stored procedure(mysql)

DELIMITER //
DROP PROCEDURE IF EXISTS GetFilteredInterfaces;
CREATE PROCEDURE GetFilteredInterfaces(IN id varchar(255))
    BEGIN
    SELECT 'GetFiltered', INTEGRATIONSERVICEID, ISTWOPHASECOMMIT, INTEGRATIONSERVICENAME, INTEGRATIONSERVICEDESC FROM integrationservice where integrationserviceid like concat ('%', id, '%');
    END //
DELIMITER ;

ISM query

{call GetFilteredInterfaces(?)} or {call GetFilteredInterfaces(#id_filter#)}

Input data structure is used to for input parameter mapping.

Output data structure is same with the above case.

• stored procedure has both input/output parameters and returns result via output parameters.

Contents of stored procedure(mysql)

DELIMITER //
DROP PROCEDURE IF EXISTS GetInterfaceWithOutput;
CREATE PROCEDURE GetInterfaceWithOutput(IN id_filter varchar(255),
        OUT out_header varchar(255),
        OUT out_id varchar(255),
        OUT out_2pc varchar(1),
        OUT out_name varchar(255),
        OUT out_desc varchar(255)
        )
    BEGIN
    SELECT 'GetWithOutput', INTEGRATIONSERVICEID, ISTWOPHASECOMMIT, INTEGRATIONSERVICENAME, INTEGRATIONSERVICEDESC
    INTO out_header, out_id, out_2pc, out_name, out_desc
    FROM integrationservice where integrationserviceid like concat ('%', id_filter, '%');
    END //
DELIMITER ;

ISM query

{call GetInterfaceWithOutput(#id_filter#, ?, ?, ?, ?, ?)} or {call GetInterfaceWithOutput(?, ?, ?, ?, ?, ?)}

Input data structure

Input data structure must contain both input and output parameters to register output parameters while invoking stored procedure.

Output data structure

Input and output data structure must have same field group.

File Service

Data structure in realtime service

When a file service is used in realtime interface, it has return data with only one field named FILE_NAME.

FILE_NAME is fixed field name in ISM. The length can be modified.

JMS Service

Web Service

EJB Service

Service mapping

Service mapping means every mapping involves output(s) of previous step(s) and input of next step. Each step stands for a service. One step may involve several output data to make its own input data.

Service mapping consists of one or more input services and one output service. Refer to File-DB Batch tutorial for the steps to define a mapping.

Data structures used for mapping are followings.

• Input service(s)
• Output data structure
• Error data structure
• Output service
• Input data structure

• Connect by name
• For file or table, ISM connects wires input field to output field with same name.
• Add default value
• For a output field, default value can be assigned. When a default value is assigned, regardless the mapping wires, the default value is set to output value.
• Add conversion function
• For a output field, conversion function can be assigned for arithmetic calculation, concatenation, trim, and etc.

Function template

Function template manages user defined custom conversion functions.

User can create his/her own function in web console.

If user code requires 3rd party libraries other than jdk like JDBC, write code in external IDE, and put compiled class to $ISM_HOME/custom directory.

Below is the bundled functions list.

Code conversion

Code mapping is used to convert meta code like values from one type to another type. It does not involve any calculation but map input value to output value.

Common

Common tab defines general explanation of code mapping definition.

Mapping

Mapping tab defines input/output value list.

Service model

Service model defines flow of an interface. Physical system is not included in service model. User defines flow with previously defined service and mapping information. Refer to Realtime Message-DB tutorial.

Common

Common tab defines basic information about a service model.

Used only in realtime and deferred services

Flow design

Flow design designs service flow from source to target(s) and how to.

Source

• Service
• Click search button() to set service

• pattern
• pattern indicates how to communicate with source system.
• Sync request
• Source will wait response in a synchronous way.
• Async request
• Source will wait response in an asynchronous way.
• Async transfer
• Source will not wait response.
• ACK request
• Source will wait response in an asynchronous way but ACK is required.
• ACK transfer
• Source will not wait response but ACK is required.

red button() means send/request. blue button() means response.

Routing

Routing has two types. One is used to send/request to target system. The other is used to respond to source or send/request to another target.

Send/Request

• Upper box is used to routing method and add or remove routing targets.
• Green button() adds sequential routing target.
• Orange button() adds concurrent routing target.
• remove button() removes last routing target.

• Lower box is used to specify mapping and routing target.
• Click search button() to choose mapping.
• Click clear button() to unset mapping.
• Click red button( and blue button() to wire routing target.
• Checkbox is default routing target. When no matching condition is found, ISM sends to this routing target.
• Left text field is routing value.
• Right text field is mapping id.
• Routing method is not All, ISM looks for matching routing target based on the value set in the routing value field.

Response

• Upper box is used to routing method, error mapping and add or remove routing targets.
• If error response returns from target system, ISM returns error response without further checking.
• If error mapping is defined, ISM transforms error response with error mapping and returns.
• Lower box is used to specify mapping and routing target.
• The layout of lower box is same with Send/Request routing box.

Routing method

Routing method has four types.

• All
• Send to all the routing targets.
• Value
• Send to the routing target which contains exactly same value with routing input.
• Function
• Send to the routing target which contains exactly same value with the result of function execution.
• Function is user defined java class. Refer to Routing function about how to implement.
• Function returns true when routing input matches condition, otherwise returns false.
• Pattern
• Send to the routing target of which routing input matches the pattern.
• Pattern means regular expression.

To determine routing target, routing input field(s) are selected. Routing input(s) are selected from output messages of previous steps. Output messages are same with service mapping.

Target

• Service
• Click search button() to set service

Pattern property is same with source.

• pattern
• Sync request
• Target will respond in a synchronous way.
• Async request
• Target will respond in an asynchronous way.
• Async transfer
• Target will not send response.
• ACK request
• Target will send response in an asynchronous way and ACK will be returned.
• ACK transfer
• Target will not send response but ACK will be returned.
• blue button() means send/Request.
• red button() means response.

Click routing button() to add routing.

Task

Task is used to do pre or post processing before or after main processing in batch and deferred services. Each task consists a group of jobs.

One interface can have only one pre task and one post task at best.

Common

Common tab defines basic information about a task.

Task

Task tab defines a set of jobs. Each job is executed sequentially according to the index from 0. If a job fails, ISM stops processing.

Job definition

Table task

Table task does not support select query. Even if you enter select query, ISM does not get query result. ISM just check whether the query is successful or not.

Telnet task

(*) To encrypt password, use encrypt.sh(bat) command. $>encrypt.sh raw_string_to_be_encrypted

ex) /sw../../bin/util>encrypt.sh solulink IUbOQ0vlgMKPF77kYaNmkw==

Stored procedure task

ISM does not use output parameters after execution. ISM just checks whether execution is successful or not.

Stored procedure call query looks like this.

{call sp_name('string_param_value', number_param_value)}

Don't forget {} at both ends of the query.

Service task

Service task is used to invoke application or service on target system. This task generates message to be used to invoke ISM realtime interface and send to realtime module. That realtime interface forwards incoming message to target system and returns result to service task. Message handler class is used to generate message for ISM realtime interface. If no message handler class is defined, the value on data property is sent as it is.

Function

• Usage

To indicate function, enclose function invocation with <!, !>. If parameter(s) exist(s) inside function invocation, ISM converts parameters then invokes function. After function is executed, ISM pass the result to handler class.

• sample

<!F(#TGT_FILE_NAME#,20,1," ")!>

• Function class

Refer to Service task function about how to implement a function.

Handler class

Handler class is used for service task and custom task. Handler does

• generates message to be sent to target system
• or execute some business logic
• and determine whether the result is success or failure

Refer to Task handler about how it is used and implemented.

Interface

Interface is final assembly of previously defined items.

• Interface type
• properties for selected type
• system

are assigned here.

After generating interface, publish to master, validate() whether it can be sent to runtime.

Common properties

Realtime service

Properties

System properties

Batch service

Properties

System properties

Deferred service

Properties

System properties