Karen,
One thing that I did for "big things" like this was to create a routine that runs on the server once a month at some off time (1am) and then send the report to whoever. Jan -----Original Message----- From: Ben Petersen <[email protected]> To: [email protected] (RBASE-L Mailing List) Date: Fri, 10 Aug 2012 09:03:23 -0700 Subject: [RBASE-L] - Re: Cursor ideas? You could set up a counter variable and have the while loop conditioned on that. reset the counter, the cursor and other variables and start again until done. That said, Mike's solution is much more elegant if it works. Working with the cursor this way only has the advantage of discovering if there is something in the environment, rather than the amount of data, that is causing the problem. On Fri, Aug 10, 2012 at 8:42 AM, <[email protected]> wrote: I was thinking of going the same place you're saying, Ben. Yes it does complete with a smaller subset. I was thinking of adding a column to the table that I can mark with an "X" after the cursor completes, have the cursor say "where flagcolumn is null" and have it quit after say 10,000 rows. Yeah, kind of a pain in that someone would have to exit RBase, come back in and start it again (would also for sure have to turn it into a permanent table). Karen In a message dated 8/10/2012 9:30:00 AM Central Daylight Time, [email protected] writes: Hi Karen, Have you tried reducing the data to ~2/3s the original to see if it will actually complete? If that works, maybe note how much can be done at one time and set the routine up to restart every X rows. That would be a pretty tacky work-a-round but it might give you a place to start. And if the routine won't actually complete using a smaller data set, then you'll know it's not the quantity of data that is the problem. Ben
