Meta dump Work Files 2 environments

 10 Replies
 0 Subscribed to this topic
 27 Subscribed to this forum
Sort:
Author
Messages
This_Guy
Veteran Member Send Private Message
Posts: 93
Veteran Member

I thought we could do a full metadumpwrk from one Productline and metaload to another. Not one by one..? How would I do a full work file dump, to then load to another env/ProductLine? 

 

 

PKS
Advanced Member Send Private Message
Posts: 23
Advanced Member
metadumpwork will dump all workfile definitions as well .
Please note to load the dependent metadata such as elements before loading the workfiles .

Appmetaload is normally used when you are dump/loading entire productline .
It does all metadata (SystemCode ,Elements ,dbdef , Programs , workdef , libdef , tokendef ,kndef, skndef, msgmnt , help text, languages )
This_Guy
Veteran Member Send Private Message
Posts: 93
Veteran Member
Thanks PKS. I did an appmetaload way back when this new PL was setup. But, see inconsistency with the work file defs. I thought I could do a metadumpwrk GLOBAL from one PL, and then load to the new PL, but the syntax is not correct. Wonder if I should just do the appmetload utility again on the new PL?
Norm
Veteran Member Send Private Message
Posts: 40
Veteran Member
When I create new product lines, I've always used the sysdump and sysload utilities, that brings over all of the various GEN tables very nicely (except menus). Not sure how that would affect your case, where you're trying to clean up inconsistencies. Might be dangerous given how much gets loaded
This_Guy
Veteran Member Send Private Message
Posts: 93
Veteran Member
Thanks Norm. I used to use sysload, but stopped based on the "new" Infor process. Either way, I was able to reconcile this by simply dumping gen wkflopts file, and importing to target - which was ultimately my issue. So, good to go!
GlendaR
Basic Member Send Private Message
Posts: 7
Basic Member
When I use metadumpwrk/metaloadwrk on file type CSV, the CSV Workfile Attributes (e.g. Field Separator, Quote Character) don't get carried over. Am I missing an option in the command or is there another metadump/metaload command that takes care of these? Any info will be much appreciated. Thanks.
jdever
Advanced Member Send Private Message
Posts: 29
Advanced Member
For CSV files:
Try rngdbdump -c product-line tableName > output.txt
and
importdb -fu new-product-line tableName output.txt
GlendaR
Basic Member Send Private Message
Posts: 7
Basic Member
Thank you jdever for the response.   However, I'm not sure if that will resolve my issue.   The solution you proposed is more for the data in the table.   The problem I'm having is when I'm copying a program which contains a workfile into another environment and the workfile is a file type CSV.   In the past, we use pgmdump/pgmload to copy over programs and all workfile definitions get copied over.   Recently, we're converting to metadump*/metaload* commands and with the metadumpwrk/metaloadwrk, which is supposed to take care of workfiles, the field definitions get copied over but not the CSV attributes.  I'm wondering if anyone else encountered this issue.  
Kwane McNeal
Veteran Member Send Private Message
Posts: 479
Veteran Member
@GlendaR, Yes, that has been broken for a while. I have opened a ticket, but I think they are still trying to fix it. I also manually set the flags in the file, but it will not load them. This one issue is why the clients I work with were forced to go back to {pgm|sys}{dump|load}
GlendaR
Basic Member Send Private Message
Posts: 7
Basic Member
Thank you Kwane for confirming the issue. I wasn't sure if I'm missing a step. I also have a ticket with Infor, still waiting for their response. This is definitely a concern, if the metload* utilities cannot fully replace the pgmdump/load utility.
JasonP
Advanced Member Send Private Message
Posts: 28
Advanced Member
Is this still a bug? What's the work around which tables do we need to dump and import to get the workdef cvs?