2 replies to this topic

[jcjones]

Native development, pro-isam file question.

I have a history file that I'm generating reports from. The file is over 2.5 million records long and has 4 key fields.

I am using SEL-ONLY on the first key field in the default logic of an LU and processing other fields in the after read no error logic.

The problem is that PROIV works on reading this file for over 25 minutes before it actually gets to the after read no error logic. The SEL-ONLY is working becuase only records with the desired value in the frist key are being read, but why does it take so long to get to reading the correct records from the file? Generally file access is very fast, and screens that display data from that file run quickly without even using the keys. Is there something else that I need to do to improve the performance of reading this file?

[Joseph Bove]

JCJones,

Is there any sort logic in place?

If there is, the function will load of the records into a dynamically generated sort (srt) file. When this happens, the sort logic executes for all of the records prior to the after read logic.

hth,

Joseph

[jcjones]

Joseph,

Thank you very much - this function was originally talking to a different source file which did have sorting, but the requirements of the report changed so it had to deal with the history file and while the sorting parameters were removed, the sort flag was still in place. So, I guess it was sorting the file by nothing, but still sorting the entire file.

I removed the sort flag and now it gets to the requested records immediately.

Certainly a "duh" moment..

Thanks,
JC Jones