|
View previous topic :: View next topic
|
| Author |
Message |
gchitra
New User
Joined: 08 Jun 2005
Posts: 36
Location: Burtonsville, MD
|
|
|
|
Hi,
We have two mainframes-East & West and files need to be transferred from East to west. After the successful NDM transmission, jobs need to be trigerred.
Could anyone plz let me know how to do this?
Also, I need a sample JCL with NDM process. I have worked only on FTP.
Thanks in advance.. |
|
| Back to top |
|
 |
superk
Global Moderator
Joined: 26 Apr 2004
Posts: 4652
Location: Raleigh, NC, USA
|
|
|
|
Well, what is your current process when the data is FTP'd?
What would you like to do? |
|
| Back to top |
|
|
gchitra
New User
Joined: 08 Jun 2005
Posts: 36
Location: Burtonsville, MD
|
|
|
|
| No FTP has been done yet. I have to do the same so that files will be available to the vendor. The only thing I know is NDM is used for FTP process. |
|
| Back to top |
|
|
gchitra
New User
Joined: 08 Jun 2005
Posts: 36
Location: Burtonsville, MD
|
|
|
|
| Once the data is FTP-ed, I need to create a trigger job to transfer the files to an sFTP system. |
|
| Back to top |
|
|
gchitra
New User
Joined: 08 Jun 2005
Posts: 36
Location: Burtonsville, MD
|
|
|
|
I just got a sample code for NDM. But didn't understand anything..
Following are the same:
NDMTOTPA PROCESS SNODE=GTEDS.TPA
STEP01 COPY FROM(PNODE DSN='....' DISP=SHR)-
TO (DSN='????(+1)'-
DISP=(NEW,CATLG,DELETE))
Could you plz tell me what it means.. |
|
| Back to top |
|
|
superk
Global Moderator
Joined: 26 Apr 2004
Posts: 4652
Location: Raleigh, NC, USA
|
|
|
|
Unfortunately, neither I nor anyone else can easily answer your question, since we do not know what your site standards would dictate. I would recommend that you contact your Connect:Direct OS.390 Administrator and ask them. These are some of the options:
- Does the target SNODE (West) use a job scheduling system? Most job scheduling systems can easily react to the event of a dataset being cataloged.
- Do both the PNODE (East) and SNODE (West) systems share a common job scheduling system? Can this system easily schedule a job to run on West when a job on East is finished?
- You can use a RUN TASK statement in your Connect:Direct OS/390 process to invoke a program on the SNODE (West) site, which can then talk to the job scheduler or perform any other task.
- You can use a RUN JOB statement in your Connect:Direct OS/390 process to invoke a job on the SNODE (West) site.
- You can use a SUBMIT PROCESS statement to submit another Connect:Direct OS/390 process on the SNODE (West) site.
I'd recommend that you obtain copies of the "Connect:Direct Process Statements Guide" and the "Connect:Direct OS/390 User's Guide" using your companie's account access code on Use [URL] BBCode for External Links. |
|
| Back to top |
|
|
superk
Global Moderator
Joined: 26 Apr 2004
Posts: 4652
Location: Raleigh, NC, USA
|
|
|
|
| gchitra wrote: |
I just got a sample code for NDM. But didn't understand anything..
Following are the same:
NDMTOTPA PROCESS SNODE=GTEDS.TPA
STEP01 COPY FROM(PNODE DSN='....' DISP=SHR)-
TO (DSN='????(+1)'-
DISP=(NEW,CATLG,DELETE))
Could you plz tell me what it means.. |
NDMTOTPA is the name of the process.
PROCESS identifies the code as being a Connect:Direct process statement.
SNODE=GTEDS.TPA identifies the Secondary Node (SNODE) that the process will communicate with. This is the node as defined in the Connect:Direct NETMAP table.
STEP01 optional step name.
COPY designates that this will be a COPY process.
FROM(PNODE DSN= identifies the source of the data, which is the Primary Node (PNODE) along with a dataset name and disposition.
TO designates the target data on the Secondary Node (SNODE) along with appropriate storage requirements (DISP, DCB, UNIT, etc.). |
|
| Back to top |
|
|
gchitra
New User
Joined: 08 Jun 2005
Posts: 36
Location: Burtonsville, MD
|
|
|
|
Thanks for the immediate replies..
As I am new to NDM, didn't understand the keywords RUNJOB & RUNTASK. But when researched a bit, saw the syntax for RUNJOB as
RUNJOB (DSN = dsn[(member)]) PNODE | SNODE
Does this mean that the DSN should have the job in the SNODE which needs to be triggered?
Today is my first day in the project and I was detailed about the requirement. so I am not sure abt the usage of CA-7. Will check with them and revert back. If available, then hope, I can specify this job as a dataset trigger, right??? |
|
| Back to top |
|
|
superk
Global Moderator
Joined: 26 Apr 2004
Posts: 4652
Location: Raleigh, NC, USA
|
|
|
|
It is common for the RUN JOB command to specify a PDS(MEMBER) that is unique to the Connect:Direct system. Rarely, if ever, is it the actual dataset that a production job would reside in.
Also, it is VERY common to use a RUN TASK to invoke the CA-7 U7SVC program to generate a dataset trigger event to CA-7. |
|
| Back to top |
|
|
gchitra
New User
Joined: 08 Jun 2005
Posts: 36
Location: Burtonsville, MD
|
|
|
|
If PDS(Memeber) rarely has the dataset where the job resides, then where do we specify the trigger job in the PNODE, while doing the NDM?
Whether RUN JOB & RUN TASK are part of the Process, that means - is it part of STEP01(refer the step mentioned above). Please clarify..
Could you plz let me know the syntax for RUN TASK? I know that Program is one of the parameters, where we specify U7SVC program. Do you have any sample step related to te same..
Thanks in advance... |
|
| Back to top |
|
|
superk
Global Moderator
Joined: 26 Apr 2004
Posts: 4652
Location: Raleigh, NC, USA
|
|
|
|
| Quote: |
stepname RUN TASK (PGM=program-name
PARM=(parameter,parameter,...)
)
SYSOPTS="parameter,parameter,..."
PNODE | SNODE
RESTART=YES | NO
|
stepname is optional.
RUN TASK is required.
PGM=program-name is required. The program runs on the node specified and has access to DD cards allocated on that node only.
PARM=(parameter,...) and SYSOPTS="parameter,..." are both optional. They specify the actual parameters to be passed to the program. The SYSOPTS parameter is used in addition to PGM when the task needs to run on UNIX.
The actual format of the parameter list that is passed to the program consists of a 2-byte field indicating the length of the parameter followed by the parameter itself. The valid data types for PARM are:
- CLn'value' specifies a data type of character with a length of 'n'. The length is optional. If not specified, the actual length of the value is used.
- XLn'value' specifies a date type of hexadecimal with a length of 'n'.
- H'value' specifies a halfword value.
- F'value' specifies a fullword value.
- PLn'value' specifies a packed value with a length of 'n'.
If no type or length is specified, then the PARM is presumed to be of type character with a length the same as the length of the value.
PNODE specifies that the program will execute on the PNODE, which is the default.
SNODE specifies that the program will execute on the SNODE. The program must exist in a LOADLIB allocated to Connect:Direct on the specified node.
RESTART=YES|NO specifies whether or not the TASK is restarted if interrupted. |
|
| Back to top |
|
|
gchitra
New User
Joined: 08 Jun 2005
Posts: 36
Location: Burtonsville, MD
|
|
|
|
Hi,
Thanks for the clarifications.
I need one more assisstance. There are 60 files on the East mainframe which need to be NDM-ed to West mainframe. After the successful tranfer, I need to setup a trigger job to send these files to mailbox repository of the vendor/sFTP system from where Vendor can access. I want to know whether it is possible to have just one jcl/job for all these files(using GDG etc) or alteast less number of jobs rather than having 1job for each file. |
|
| Back to top |
|
|
gchitra
New User
Joined: 08 Jun 2005
Posts: 36
Location: Burtonsville, MD
|
|
|
|
What does (-0) indicate in the GDG file
abc.def.xyz(-0)?
Is this same as specifying (0), current generation? If so, then why do we specify "-"? |
|
| Back to top |
|
|
|
|
