> I'm now wondering if some error is occurring that, for some reason, is silently ending the routine. I'm building what may be a very long SQL INSERT statement for each line in the CSV file that I'm reading; could I be hitting some upper limit for the length of the SQL code? I'd think that an error would be presented in this case, but maybe I have to do something explicitly to force all errors to display? Even warnings? > > Another thing I've noticed is that the "timeout" (I'm not even certain the problem IS a timeout any longer, hence the quotation marks) doesn't happen at the same record every time. That's why I thought it was a timeout problem at first, and assumed that the varying load on the server would account for the different record numbers processed. If I were hitting some problem with the SQL statement, I'd expect it to stop at the same record every time. Or is that misguided thinking, too? 1) When you say, "doesn't happen at the same record every time" are you using the same dataset and speaking about the same line number? Or are you using different datasets and noticing that the line number varies? If it's the same dataset, it sounds like "fun" -- as in "a pain in the assets". 2) I'm writing something similar; letting a user upload a CSV file via a webpage, then creating an SQL query with many records. So now I'll be watching your thread. For debugging purposes, create your SQL statement and print it out on the webpage (or save it somewhere -- maybe a file). Don't have your webpage script execute the query. Then see if you get the complete query you expect. Then copy that query into a database tool like phpmyadmin and see if you get errors when executing the query. _________________________________________________________________ Peek-a-boo FREE Tricks & Treats for You! http://www.reallivemoms.com?ocid=TXT_TAGHM&loc=us
