Clarion accessing Postgres

odbc
Tags: #<Tag:0x00007f60c0c40cc0>

#1

Hi, everybody, we are migrating from Oracle to Postgresql, and having troubles on using ODBC driver. The template Browse is very slow to listing data. We already saw something about the use of “Prop:SQLOrder”, to implement a “LIMIT” on fetching data, but having trouble to make this works, once the scrolling of the data is disordered, and some data repeats. We are using the legacy template, not the ABC.
There’s anybody that can help us? Any help is valid.


#2

Not sure about the legacy templates, but with ABC & PostgreSQL I have found it surprisingly fast.
All browses, file combos, file drops are file loaded, usually with a range/filter sufficient to keep the number of records reasonable.

Defining any more complicated views on the backend definitely helps and will give you the opportunity to test and fine-tune the speed.

Though not said to be important by some, when I first moved to PG using ODBC, I found setting NESTING=True in the driver string to be beneficial. I added this to my templates and now never have to think about it as a setting. It is definitely something you should try.

Cheers,
Douglas


#3

totalhip, thanks for your attention. I tried to put the “NESTING=TRUE” on the envolved tables, but don’t looks better. Do you believe that creating views on the backend can resolves? Thanks, one more time.


#4

I think only the most complex views might need to go the backend.
I would start with a single table and get that going with speed first.
For ~10K records, the data certainly comes back to my App <100ms.
It actually is only a bit slower than with PgAdmin.


#5

Have a look at your connect string and add UseDeclareFetch=1 to it.
That can make a huge difference is speed on some browses