tpayne
Reputable Poster
We upgraded our World system from A8.1 cum5 to cum6 yesterday, and it went reasonably smooth with one exception, the GL Posting jobs (J09800) started to take 4+ hours to run instead of minutes.
It took us a few hours to find the cause, but I was curious to whether any other companies have the strange anomoly as we do with F0911LQ... I'm not sure if this is just an A8.1 file or if it exists also in A7.3...
This logical has selected all records where GLPOST = 'P' and GLALT4 <> 'P', ie: all posted transactions where one of the Alternate Posting flags is not showing Posted.
In our production system prior to the upgrade, as well as all the other copies of the file we have, including the objects from JDE, the index contains zero records, however in production it now shows 78 million records. Either use DSPFD F0911LQ or WRKF F0911LQ and option 8 - the number of entries in the Index is shown near the end of the list. If the file is corrupted, the number of index entries should be showing as zero.
It looks as if this file was shipped corrupted all along, the source member and object dates dating back to 1999, however doing a CRTDUBOBJ as part of the upgrade the index now "correctly" includes the records that it should. Of course that now means the first posting job is running programs P51801 and P51802 as part of J09800 and taking hours to run, processing through 78 million records...
It took us a few hours to find the cause, but I was curious to whether any other companies have the strange anomoly as we do with F0911LQ... I'm not sure if this is just an A8.1 file or if it exists also in A7.3...
This logical has selected all records where GLPOST = 'P' and GLALT4 <> 'P', ie: all posted transactions where one of the Alternate Posting flags is not showing Posted.
In our production system prior to the upgrade, as well as all the other copies of the file we have, including the objects from JDE, the index contains zero records, however in production it now shows 78 million records. Either use DSPFD F0911LQ or WRKF F0911LQ and option 8 - the number of entries in the Index is shown near the end of the list. If the file is corrupted, the number of index entries should be showing as zero.
It looks as if this file was shipped corrupted all along, the source member and object dates dating back to 1999, however doing a CRTDUBOBJ as part of the upgrade the index now "correctly" includes the records that it should. Of course that now means the first posting job is running programs P51801 and P51802 as part of J09800 and taking hours to run, processing through 78 million records...