About transaction processing controls
Transaction communications are independently controlled processes between the client, server, and host platforms. Control functions allow you to manage transaction processes.
Data collection server control
At the data collection server, you can start and stop the data transmission between the TranConnect Host Agent and the TranConnect polling program. The TranConnect Host Agent retrieves the transaction records from the Transaction Log table and sends them to the polling program on the host computer. Stopping this agent temporarily stops data transmission from the data collection server to the host. Workers on the shop floor can continue to collect data, but none will be sent to the host. You use the stop/start agent option in System Monitor.
Host control
At the host, you can control transaction processing by starting and stopping the Polling program and starting and stopping the pipeline, which consists of the Formatter and Processor programs. The STRPOLLSRV command starts the job and the WRKSBS command is used to end the job.
Stopping polling temporarily stops data transmission, but the pipeline continues to process transaction records. You control the pipeline through the Start/Stop Transaction Processing (SST) function in the TranConnect program. Stopping the pipeline temporarily stops transaction processing, polling, and syncing.
The WHT function is a control used for processing held transactions. The WHT function lets you release the hold on the transaction record, edit the record, or delete the record.
Platform Manager transaction controls
Platform manager provides polling, pipeline, subsystem, and batch routine transaction controls.
For more information, refer to one of the following topics.
About polling control
The polling (SocketPoll) program uses TCP/IP protocol communications. You can start polling by using use the STRPOLLSRV command to start polling. Use stop polling by using the WRKSBS command to stop polling.
For more information, refer to Start polling or Stop polling.
Start polling
To start polling, complete the following steps.
-
Start the DCLINKSBS susbsystem.Step InformationNote:If the subystem is not active, the SocketPoll server job waits in the job queue until the system starts.
-
On the command line, enter STRPOLLSVR and press F4.
-
For SocketPoll Server Job Name, enter the job name for the SocketPoll server job.Step InformationNote:The job name must match the entry in the Data Collection Server (DCS).
-
Verify the default values of the following fields and make any applicable changes.
-
IP Port Number
-
SocketPoll Server Program Lib (Library)
-
Job Description
-
Job Description Library
-
Job Queue
-
Job Queue Library
-
-
Press or select Enter.
-
On a command line, enter WRKSBS.
-
For Opt, in the DCLINKBS subsystem field, enter 8.Step InformationNote:You can also use the SST menu exit to review the status of the SocketPoll server job.
-
Ensure the status for all started jobs in the PC-Polling column in SST is *ACTIVEsoc.
-
Close the form.
Stop polling
You can use an end job command to stop polling.
To stop polling, complete the following steps.
-
On a command line, enter WRKSBS and press F4.
-
For Opt, next to the DCLINKSBS subsystem, enter 8 and press or select Enter.
-
Enter 4 and press or select Enter to stop the job.
About pipeline control
You can verify the status of the pipeline in the Start/Stop Transaction Processing form or the subsystem.
When you stop the TranConnect Host Agent or pipeline, data is still collected from the devices and queued on the data collection server until you restart polling. You must stop the pipeline before performing setup or routine maintenance functions.
You can use SST functions to perform the following actions.
-
Start or stop the pipeline.
-
End PC collections to stop polling.
-
Force the pipeline off to immediately halt, and end the subsystem.
For more information, refer to Start or stop the pipeline.
Start or stop the pipeline
Starting and stopping the pipeline takes a few seconds.
Prerequisites:
-
The pipeline cannot be started unless the current status of the pipeline is *OFF.
-
The pipeline cannot be stopped unless the current status of the pipeline is *ACTIVE.
To start or stop the pipeline, complete the following steps.
-
On a menu selection line, enter SST and press Enter or select OK.Expected ResultResult: The Start/Stop Transaction Processing form displays.
-
Do one of the following steps.
-
Select Start Pipeline to start all pipeline jobs for the data collection servers assigned to the same data queue.
-
Select Stop Pipeline to stop the DCLPOLL job for the data collection servers assigned to the same data queue.
-
Select Force All Off to immediately stop all pipeline jobs, end the subsystem, and clear the job queues in the subsystem.
-
Press F16 or Shift+F4 to start all pipeline jobs for all data queues.
-
Press F18 or Shift+F6 to stop all pipeline jobs for all data queues.
-
Verify the status of a subsystem
You can verify the status of a subsystem using the Work with Subsystems (WRKSBS) or Work with Active Jobs (WRKACTJOB) forms. The polling, pipeline, and synchronization jobs for transaction processing run in the DCLINKSBS subsystem.
To verify the status of a subsystem, complete the following steps.
-
On a command line, enter WRKSBS and press Enter.Expected ResultResult: The Work with Subsystems prompt displays.
-
Navigate to the subsystem you want to review
-
In the field next to the subsystem you want to review, for OPT, enter 8 and press Enter.Expected ResultResult: The Work with Subsystem Jobs form displays and shows the active jobs in the subsystem.
About batch routines
Attended or unattended batch routines can start or stop transaction processing as part of nightly batch schedules. Batch routines can start and end polling, pipeline, and synchronization jobs.
You can run a program that contains the commands for starting the pipeline, stopping the pipeline, and ending all pipeline jobs and the subsystem in which they run. Running an attended batch routine requires an operator to enter the batch command. If you are running a batch routine as a nightly procedure, coordinate the schedule with the operator in charge of the processing server, that runs the TranConnect Host Agent, TranConnect Synchronization Agent, or both. These agents must be stopped through the System Monitor that runs on the processing server.
When using batch routines, it is important to keep the following things in mind.
-
Ending or stopping the pipeline in batch must be done cautiously, as this function stops the pipeline jobs even if there are unprocessed transactions.
-
Do not attempt to start or stop the pipeline while the status is either *STARTED or *SHUTDOWN. Starting or stopping a pipeline can be a lengthy process. Attempting a start while in *SHUTDOWN status will fail because the start process will not submit the job. Likewise, attempting a stop while in *STARTED status will fail because the stop process will not submit the job.
Run an attended batch routine to start the pipeline
This procedure starts all pipeline jobs that have a status of *OFF. This command also starts the subsystem.
To run an attended batch routine to start the pipeline, complete the following steps.
-
On a command line, enter CALL PGM(DC10SSTSA) and press Enter.Expected ResultResult: If it is not running, the subsystem starts, and all formatter and processor jobs for the listed data collection servers that have a status of *OFF start. Any transactions left in the data queues when the pipeline last ended are processed.
-
Optional. If the TranConnect Host Agent was stopped, restart it from the System Monitor.
Run an attended batch routine to end the pipeline
The end process stops the active jobs (polling, pipeline, and synchronization) and ends the subsystem in which they run.
If you are synchronizing records, ensure TranConnect Synchronization Agent has stopped and all records are synchronized before running the batch routine.
To run an attended batch routine to end the pipeline, complete the following steps.
-
On a command line, enter CALL PGM(DC10SSTEA) and press Enter.Expected ResultResult: Transactions in data queues remain in the data queues and process when the pipeline starts again.
About running unattended batch routines
You can run batch routines using the AS/400 Job Scheduler function. For example, you can schedule a batch routine to start all jobs in a particular subsystem before the first shift of the day and another batch routine to end all jobs in the same subsystem after the last shift.
Before implementing these functions into your schedule, it is recommended that you consult with your Advanced Inventory consultant. Specific user profiles and library lists must be used to start the pipeline. Other jobs or special exit data queue programs may need special consideration.
The following commands are examples of how to automate the stopping and ending jobs using the job schedule function of AS/400.
Run unattended batch routines in AS/400 job scheduler
You can enter commands on a command line or in a prompt form. The following examples show the command line for starting the jobs (polling, pipeline, and synchronization) in the subsystem at 5:00 a.m. and ending them at 11:00 p.m. everyday.
You must enter three separate commands for starting the pipeline, polling, and synchronization jobs.
To run unattended batch routines in AS/400 job scheduler, complete the following steps.
Note:The commands in the samples below use the default library names. If your library names are different, replace the default library names with the name of your libraries. Also note that the commands for starting the jobs for polling and synchronization contain the default IPPort values. Replace these values with your IP Port addresses.
-
On the command line, enter a command to start the pipeline.ExampleExample:
ADDJOBSCDE JOB(STRPIPELIN) CMD(CALL PGM(DC10SSTSA) + FRQ(*WEEKLY) SCDDATE(*NONE) SCDDAY(*ALL) SCDTIME(050000) + JOBD(DCLINKOBJ/DCLINKJBD1) JOBQ(QBATCH) USER(DCLINK) + TEXT('Start All dcLINK "Pipelines"') -
On the command line, enter a command to start the polling job.ExampleExample:
ADDJOBSCDE JOB(STRTPOLL) CMD(STRPOLLSVR JOBNAME(SOCPOLL) IPPORT(2000) + PGMLIB(*LIBL) JOBD(DCLINKJBD2) JOBDLIB(*LIBL) JOBQ(DCLINKJBQ2) + JOBQLIB(*LIBL) FRQ(*WEEKLY) SCDDATE(*NONE) + SCDDAY(*ALL) SCDTIME(050000) JOBQ(QBATCH) USER(DCLINK) + TEXT('StartdcLINK "Polling"') -
On the command line, enter a command to start the synchronization job (SocketSync).ExampleExample:
ADDJOBSCDE JOB(STRTSYNC) CMD(STRSYNCSVR JOBNAME(SYNCSVR) IPPORT(2100) + PGMLIB(*LIBL) JOBD(*SYSDTAARA) JOBDLIB(*SYSDTAARA) JOBQ(DCLINKJBQ2) + JOBQLIB(*LIBL) USRSPCLIB(*UDC) PURGEJOB(*NO)) FRQ(*WEEKLY) SCDDATE(*NONE) + SCDDAY(*ALL) SCDTIME(050000) JOBQ(QBATCH) USER(DCLINK) + TEXT('START dcLINK "Sync"') -
On the command line, enter a command to end the pipeline.Step InformationNote:You only need to enter the command to end the pipeline because the program ends the pipeline jobs and the subsystem in which they run.ExampleExample:
ADDJOBSCDE JOB(ENDPIPELIN) CMD(CALL PGM(DC10SSTEA)) + FRQ(*WEEKLY) SCDDATE(*NONE) SCDDAY(*ALL) + SCDTIME(230000) JOBD(DCLINKOBJ/DCLINKJBD1) JOBQ(QBATCH) USER(DCLINK) + TEXT('End All dcLINK "Pipelines"')
Loading...
There was a problem loading this topic