Recent content by jsveiga

Thank you Tamhas, Do you have any specific concern I should watch out for about the DROP TABLE / ADD TABLE method - apart from the non-reducing size (150MB, not a big deal)? I mean, the .df generated from the production database contains all the db structure, and should have reproduced it back...

Hi, they were the ones who first suggested a clean install, but I managed to talk with other person there (more technical this time) and he suggested the use of the "\empty" db files I mentioned before - actually a "\presetup" directory with some extra initialization made (and find out the...

Thank you again :-) I've found that the original SW installation has an "\empty" directory with the original db files (structures without data). I could then find out what the patch level was for the original install, and apply all the delta .df's to get to the current version (so no need to...

Thank you Tom, I did check the "POSITION for .r code compatibility" option when creating the .df, but I still get "CRC mismatch, try recompiling" when running the app. I cannot recompile because I don't have access to the sources. It is an ERP system with a huge amount of .r programs. I am...

Hello, I need to create a copy of an existing OpenEdge 10.1A database, but without its data - only replicate all the table structures, keeping .r compatibility. Or (final result would be the same) just a way to purge all records from all tables of the database. I saw a suggestion on...

Hello, Is there a reason why proshut (_mprshut) sometimes shows, and sometimes does not show the client host name? For example, here is the output of a _mprshut -C list: usr pid time of login user id tty Limbo? 13 4292 Mon Jan 04 17:09:07 2010 BIW...

... and remember that up to Aug/19 14:30 it worked fine with my query returning 29500 rows and the fetch array size was the default 50!

Got a better solution (although still no explanation), WITHOUT the sleeps. I reduced my "chunk" size to 20 (i.e., now my big query which returned 25900 items runs divided in part number ranges of max 20 codes - it was 50 up to last tests). With this I was able (up to now) to run 9 times...

Thank you Tom (sorry it takes me so long to answer; for every change I have to let the import run sometimes to see if the result is consistent) I just did this test: 1 - removed the "un-optimization" of closing/destroying/creating/opening the ODBC connection every time (which works fine). This...

Thank you very much Netrista, I will probably try the driver again soon, and will contact support for the correct trial license. The problem finding the crash in the server error log is that there is no crash :-/ Note that on most of the cases, I just get no answer for the query, and my...

Hi Netrista, (btw, one of the things I tried was to install the OpenLink Progress ODBC driver 6.1 with a trial license, but when defining the DSN it said "OS not licensed" for the OpenEdge 9.1 driver - maybe because I tried in Vista and in Win2003 64bits?, and says there is a missing dll for...

Thank you Casper, For the one time I managed to make the app run in the server yesterday but got horrible performance I believe I got the explanation: I had left ODBC Tracing enabled in the server. For the original problem (hanging on the big query), the Tracing may have given me a clue: I...

I got the app running on the server once (it returned the "big" SELECT). To put things in numbers: The "big" SELECT returns about 25900 rows, then I have to fetch more info (no "LIMIT" or "TOP" on 10.1A...) from other tables, with other SELECTs, one for each of these 25900 rows. In my...

Hello, We are running OpenEdge 10.1A-02 (I know, old version, came with our ERP), and I have an application which periodically (every 10 min) pulls data from the Progress DB (read only ODBC connection) to a MySQL DB. The ODBC driver came with OpenEdge, 05.10.0039. The DBs and the app runs on...